Skip to main content

Full text of "Celtic art in pagan and Christian times"

See other formats







17 i 3 





/ Preface . . . ... 


>^ The Continental Celts and How they Came to 

Britain . . ... 







Pagan Celtic Art in the Early Iron Age . . 90 

Pagan Celtic Art in the Bronze Age . . . 22 

Pagan Celtic Art in the Early Iron Age . . 61 

Pagan Celtic Art in the Early Iron Age . .129 


Celtic Art of the Christian Period . . . 162 


Celtic Art of the Christian Period . . . 232 

Celtic Art of the Christian Period . . 254 





The Hallstatt Sword . . . ... 8 





Ornament on bronze sword-sheeith from La Tene 

Gaulish hehnet of bronze from Gorg-e-Meillet 

Chevron patterns of the Bronze Ag-e 

Bronze spear-heads ornamented with rows of dots 

Gold lunula from Killarney . . . . 

Long-itudinal section of chamber and passage of Tumulus at Ne\ 

grange, Co. Meath ..... 
Plan of chamber and passage of Tumulus at Ncwgrange, Co. Meath 47 
Spiral ornament at Newgrange, Co. Meath . . . . 48 

Slab with spiral ornament outside entrance to passage of Tumulus 

at Newgrange, Co. Meath . . ... 49 

Spiral ornament on bronze axe-head from Denmark . . -51 

Bronze axe-head with spiral ornament from Sweden , . . 52 

Winding band (curved swastika), sculptured on rock near Ilkley, 

Yorkshire . . . • ... 58 

Cup-and-ring sculptures on rock at Ilkley, Yorkshire . • • 59 

Bronze sword-sheaths from Hunsbury . . • • 97 

Silver-gilt fibular found in Northumberland . . . . 104 

Bronze beaded torque from Lochar Moss, Dumfriesshire . . 112 

Late-Celtic bronze spoon from Brickhill Lane, London . .119 

Late-Celtic bronze spoon from Crosby Ravensworth, Westmoreland 119 
Late-Celtic spoon, one of a pair, from Weston, near Bath . .120 
Late-Celtic urns from Shoebury, Essex . . . . 123 

Late-Celtic bronze mirror from Trelan Bahow, Cornwall . . 131 

Bronze sword-sheath from Hunsbury . . . . 132 

Late-Celtic pottery from the Glastonbury Lake Village . . 142 

Late-Celtic wooden tub found at the Glastonbury Lake Village . 147 
A 2 vii 



Handles of pair of Late-Celtic spoons from Weston, near Bath 
Engraved bone object from Slieve-na-Caillighe, Co. Meath 
Fibula of bronze-g-ilt from ^Esica . . . . 

Collar from Broig'hter, Limavady, Co. Londonderry . 

Spiral ornament in illuminated MS. copied from repousse metahvork 154 

Shading- of parallel lines . . . . . . 156 

Cross-hatching placed diagonally . . . . . 156 

Cross-hatching placed diagonally, with dots . ... 157 

Cross-hatching of double lines placed diag-onally . . • I57 

Chequerwork g-rass-matting shading- . . . . 157 

Eng-ine-turned shading- . . . ... 158 

Dotted shading . . . . ... 158 

Swastika desig-n on shield from the Thames . ... 159 

Engraved ornament found at the Glastonbury Lake villag-e . . 161 

Handles of bronze bowl found at Barlaston, Staffordshire . . 166 

Handle of bronze bowl from Chesterton-on-Fossway, Warwickshire 167 

Handle of bronze bowl from Chesterton-on-Fossway, Warwickshire 168 

Spiral ornament from the Book of Durrow . ... 169 

Cross-slab from Pen-Arthur, Pembrokeshire . . . . 181 

Erect cross-slab at St. Madoes, Perthshire . ... 183 

Cross at Penmon, Anglesey . . ... 185 

Great wheel-cross of Conbelin at Margam Abbey, Glamorganshire 187 

Cross at Neuadd Siarman, near Builth, Brecknockshire . . 189 

Cross at Nevern, Pembrokeshire . . ... 191 

Pin-brooch from Clonmacnois, King-'s Co, . . . . 221 

Pierced marble screen at Ravenna . . ... 245 

Regular plaitwork without any break . ... 259 

Method of making- breaks in plaitwork . ... 259 

Regular plaitwork, with one vertical break and one horizontal break 260 

Six-cord plait, with horizontal breaks at regular intervals . . 260 

Cross-shaft at Golden Grove, Carm^irthenshire . . . 261 

Cross-shaft at Llantwit Major, Glamorganshire . . . 261 

Eight-cord plaits, with cruciform breaks . . . 262, 263 

Six-cord plait, with cruciform breaks . . ^ . 264 

Ten-cord plait, with cruciform breaks . i . . 264 




Knots derived from a three-cord plait . ... 264 

Diag-rams of knotwork ..... 265, 266 

Method of deriving- knots Nos. 3 and 6 from a four-cord plait . . 267 

Knots Nos. 4, 5, 7, and 8 derived from a four-cord plait . . 268 

Knot No. I, derived from either a three-cord or a six-cord plait . 268 

Knots 3 and 4, derived from a six-cord plait . ... 269 

Evolution of knot No. 1 from a six-cord plait . . . 269 

Evolution of knot No. 7 from an eight-cord plait . . 271, 272 

Diagrams of knotwork . . . ... 273 

Knotwork from Ramsbury, Wilts, and Nigg, Ross-shire . . 274 

Circular knotwork from Tarbet, Ross-shire . ... 276 

Circular knotwork from Monasterboice, Co. Louth . . . 276 

Triangular knotwork from Ulbster, Caithness . . . 277 

Triangular knotwork from Dunfallandy, Perthshire . . . 278 

Key-patterns . . . . . . 280, 282 

Shaft of cross of Eiudon at Golden Grove, Carmarthenshire . . 283 

Erect cross-slab at Rosemarkie, Ross-shire . ... 284 

Methods of connecting- spirals . . ... 285 

Tree key-pattern, Meigle, Perthshire . ... 286 

Key-pattern border from the Book of Kells . ... 287 

Method of connecting spirals . . ... 288 









Grave of a Gaulish warrior at Sesto Calende, 

Italy . . ... 

Bronze fibula; of La T^ne type from the ceme- 
teries of the Marne . ... 
Bronze armlet of the La Tene period from 

Germany . . ... 

Bronze armlet of the La T<^ne period from Long-i- 

rod (Viiud) . . ... 

Bronze armlet of the La Tene period from the 

cemeteries of the Marne 
Gaulish helmet of bronze from Berru (Marne) . 
Cinerary urn of Bronze Age from Lake, Wilts, 

now in the British Museum. Height i ft. 3I ins. 
Bronze Age urn of "Incense Cup" type from 

Aldbourne, Wilts, now in the British Museum. 

Height sh i<i^- • ... 

Bronze Age urn of "Food Vessel" type from 

Alwinton, Northumberland, now in the British 

Museum. Height 5 ins. 
Bronze Ag'e urn of " Drinking Cup" type from 

Lcikenheath, Suffolk, now in the British 

Museum. Height 7^ ins. 
Spiral ornament on stone ball from Towie, -n^ 

Aberdeenshire, now in the Edinburgh Museum. 

Scale T linear . . ... 

Winding- band curved swastika on sword-hilt 

from Denmark . ... 

Bronze sword-hilt with winding- band pattern 

from Denmark . ... 

Bronze sword-hilt with spiral ornament from 

Denmark . . ... 

Bronze mirror from Birdlip, Gloucestershire, now 

in the Gloucester Museum. R. W. Dugdale, 

photo . . ... 

- To face page 8 









IX. Iron dagger with bronze hilt and sheath from 
the River Witham . ... 

X. Bronze harness rings from Polden Hill, Somerset- 
shire, now in the British Museum. Scale 
f linear . . ... 

Late-Celtic bronze fibula from Walmer, Kent, 
now in the British Museum. Scale f linear . 

Late-Celtic fibula from Ireland, now in the 
Museum of the Royal Irish Academy, Dublin 

Enamelled bronze fibula from Risingham, North- 
umberland, now in the Newcastle Museum 

Bronze fibula from Water Eaton, Oxon, now in 
the British Museum. Scale ^ linear 

Bronze fibula from Clogher, Co. Tyrone, now 
in the British Museum. Scale x linear . 

Late-Celtic bronze fibula (locality unknown), 
now in the British Museum. Scale i linear . 

S-shaped enamel bronze fibula (locality unknown), 
now in the British Museum. Scale j linear . 

S-shaped fibula of enamelled bronze from Nor- 
ton, E. Riding of Yorkshire, now in the British 
Museum. Scale { linear 

Bronze fibula from Polden Hill, Somersetshire, ^ 
now in the British Museum, side view. Scale 
f linear . . ... 

Bronze fibula from Polden Hill, Somersetshire, 
now in the British Museum, front view. Scale 
f linear . . . , 

Bronze fibula from River Churn, now in the 
British Museum, front view. Scale i linear . 

Bronze fibula from River Churn, now in the 
British Museum, side view. Scale { linear 

' Bronze hook and disc ornament from Ireland, 

now in the Dublin Museum 

Bronze pin enamelled from Danes' Graves near 

XIV. } Driffield, Yorkshire . ... 

Bronze disc fibula with Late-Celtic ornament 

from Silchester, now at Strathfieldsaye House. 

S. Victor White, of Reading, photo. 

To face page 92 

















Bronze beaded torque from Mowroad, near^ 

Rochdale. Scale f linear 
Bronze collar from Wraxhall, now in the 

Bristol Museum . ... 

Bronze armlets from the Culbin Sands, now 

at Altyre, near Forres, N. B. . 
Late-Celtic bronze mirror in the Mayer 

Museum, Liverpool (locality unknown) 
Late-Celtic pottery from Hunsbury, now in 

the Northampton Museum 
Late-Celtic pottery from Yarnton, Oxford-^ 

shire, now in the British Museum. Scale 

f linear . . ... 

Late-Celtic pottery from Kent's Cavern near 
Torquay, Devonshire, now in the British 
Museum. Scale | linear 

Granite monolith with Late-Celtic sculpture 
at Turoe, Co. Galway. Height of stone 
4 ft. Reproduced from a photograph by 
Mr. A. McGoog-an illustrating- Mr. George 
Coffey's paper in the Proceedings of the 
Royal Irish Academy 

Cruciform harness mounting of bronze ena- 
melled (locality unknown), now in the 
British Museum. Scale \ linear 

Bronze enamelled harness mounting from 
Polden Hill, Somersetshire, now in the 
British Museum. Scale % linear 

Upper part of bronze sword -sheath from 
Lisnacroghera Co., now in the British 
Museum . . ... 

Lower part of bronze sword -sheath from 
Lisnacroghera Co., now in the British 
Museum . . ... 

Handle of Late - Celtic bronze tankard, . 
Trawsfynydd, Merionethshire, now in the 
Mayer Museum, Liverpool . . . 

Bridle bit of bronze enamelled, from Rise near I 
Hull, now in the British Museum. Scale 
i linear . . ... 

-To face page no 










XXIV. Detail of ornament on Late-Celtic bronze 
shield from the Thames at Battersea, 
now in the British Museum . 

Circular disc of bronze with repouss^ 
ornament from Ireland, now in the British 
Museum . . ... 

Bronze enamelled harness-ring- from West- 

Ihall, Suffolk, now in the British Museum. 
Scale I linear . ... 

/ Cast of metal object (locality unknown) ^ 

from the Albert Way Collection, now in 
I the Museum of the Society of Antiquaries, 
XXVI. ! Burling-ton House . . ' ' f 

Cover of the Stowe Missal, in the Museum 
of the Royal Irish Academy, Dublin 
V (a.d. 1023 to 1052) . ... 

To face page 152 






XXXI. ^ 

Erect cross slab in Aberlemno Churchyard, 
Forfarshire. John Patrick, of Edinburgh, 
photo. . . ... 

Bronze bell with engraved ornament from 
Loug-h Lene Castle, Co. Westmeath, in 
the Museum of the Royal Irish Academy 

The shrine of the bell of St. Patrick's Will, 
in the Museum of the Royal Irish 
Academy, Dublin (a.d. 1091 to 1105) . 

Head of the Lismore crozier at Lismore 
Castle, Co. Waterford (a.d. 1090 to 1113) 

Celtic quadrang-ular bell of bronze with 
zoomorphic handles from Llang-wynodl 
Church, Carnarvonshire, now in the pos- 
session of Corbet Yale-Jones Parry, Esq., 
of Madryn Castle, Pwllheli. Mr. Morgan 
Evans, of Pwllheli, photo. . 

Bronze reliquary from Lower Lough Erne, 
now in the possession of T. Plunkett, 
Esq. , of Enniskillen. 7 ins. long- by 5^ ins. 
high by 3^ ins. wide. R. Welch, of Bel- 
fast, photo. . . . . 




,, 206 














Bronze fibula with plaitwork and Late- 
Celtic ornament from the Ardakillen 
Crannog-, near Strokestown, Co. Ros- 
common, now in the Museum of the 
Royal Irish Academy, Dublin 

Detail of ornament on the underside of 
the foot of the Ardag-h chalice, in the 
Museum of the Royal Irish Academy, 
Dublin . . ... 

Silver penannular brooch from Ireland, 
now in the British Museum. Scale 
5 linear . ... 

Biskra woman wearing- a pair of penannu- 
lar brooches, the ends of the pins point- 
ing- upwards . ... 

Details of ornaments on the Tara brooch 

Details of ornaments on the Tara brooch 

Details of ornament on the Tara brooch, 
in the Museum of the Royal Irish 
Academy, Dublin 

Circular knotwork on slab in church of 
Sta. Sabina, Rome 

Doorway of the chapel of S. Zeno in the 
church of S. Prassede, showing- broken 
plaitwork on jambs (a.d. 772 to 795) . 

Key pattern, S. Apollinare Nuovo, Ravenna ^ 

Vine scrolls, S. Apollinare Nuovo, Ravenna / 

Plaitwork of Romano-British pavement 
at Lydney Park, Gloucestershire . 

Plaitwork in ciborium in the church of 
San Clemente, Rome (fifth century) 

Erect cross slab from Collieburn, Suther- 
land, now in the Dunrobin Museum 

Detail of ornament on erect cross slab at 
Nigg-, Ross-shire . ... 

Four men placed swastika fashion on 
recumbent pavement at Meig-le, Perth- 
shire. Scale i linear 

Spiral ornament on frag-ment of sculptured 
stone from Tarbet, Ross-shire, now in 
the Edinburg-h Museum 

Detail of ornament on erect cross slab at 
Nig-g-, Ross-shire 

To face page 216 








THIS work is an attempt — whether successful or 
not the critic must decide — to give a concise 
summary of the facts at present available for 
forming a theory as to the origin and development of 
Celtic art in Great Britain and Ireland. By Celtic art 
is meant the art of the peoples in Europe who spoke 
the Celtic language, but it must always be borne in 
mind that although linguistically they were Celts, yet 
racially they were of mixed Celtic and Iberian blood, 
so that their art was possibly quite as much Iberian 
as Celtic. It is only since the epoch-making discoveries 
of Schliemann in Greece, of Flinders Petrie in Egypt, 
and of Arthur Evans in Crete that it has been possible 
in a satisfactory manner to connect the culture of 
Britain in the Bronze Age with the corresponding 
culture on the Continent. It is now quite clear that 
certain characteristic decorative motives, such as the 
divergent spiral, are of foreign origin instead of having 
been invented in Ireland, as was at one time believed. 
Other discoveries made in England, more especially 
those at Aylesford, Glastonbury, Mount Caburn, and 


Hunsbury, have thrown an entirely new light on the 
archaeology of this country by showing that the Early 
Iron Age began here two or three centuries at least 
before the Roman occupation. Lastly, the explorations 
made by Continental antiquaries at Hallstatt in Austria, 
La Tene in Switzerland, and in the Gaulish cemeteries 
of the Marne district in France, point to the sources 
of the culture to which the late Sir Wollaston Franks 
gave the name '' Late-Celtic." 

Celtic art naturally divides itself into two distinct 
periods, the Pagan and the Christian. With regard 
to the latter, the remains have been so fully investigated 
that it is hardly probable any new facts will be brought 
to light which will seriously alter the conclusions now 
arrived at. With regard to the Pagan period the case 
is altogether different, as most of the finds hitherto 
made have been due to accident, and until the large 
number of inhabited and fortified sites belonging to 
this period are systematically excavated our knowledge 
must necessarily remain incomplete. 

I have endeavoured to give in the footnotes all the 
sources whence my information has been obtained, but 
I should like more especially to acknowledge my in- 
debtedness to A. Bertrand and S. Reinach's Les Celtes 
dans les Vallees dii P6 et du Danube; J. Anderson's 
Scotlaitd in Pagan Times and Scotland in Christian 
Times; Arthur Evans' papers on the Aylesford, ^sica, 
and Limavady finds in the Archceologia ; and George 



Coffey's papers on the ornament of the Bronze Age, 
Newgrange, etc., in the Journal of the Royal Society 
of Antiquaries of Ireland, and in the Transactions and 
Proceedings of the Royal Irish Academy. 

The theory of the evolution of Celtic knotwork out 
of plaitwork (as explained on pages 257 to 278) is 
entirely original, and, simple as it appears when ex- 
plained, took me quite twenty years to think out 
whilst classifying the patterns that occur on the early 
Christian monuments of Scotland, England, Wales, 
and Ireland, nearly all of which I have examined per- 

No illustrations are given of the pages of the Celtic 
illuminated MSS. on account of the difficulty of making 
satisfactory reproductions of them on a small scale. 
I have thought it better to refer the reader either 
to the MSS. themselves or to the Publications of the 
Palceographical Society and Professor J. O. Westwood's 
Miniatures of the Anglo-Saxon and Irish Manuscripts. 

A large number of photographs of Late-Celtic metal- 
work in the British Museum have been specially taken 
for this work by Mr. H. Oldland, with the kind per- 
mission of Mr. C. H. Read, f.s.a. I am indebted to 
the Rev. Canon W. Bazeley for obtaining a photograph 
of the Birdlip mirror in the Gloucester Museum, and 
to Mr. George for the loan of Sir H, Dryden's drawings 
of the Hunsbury sword-sheath in the Northampton 
Museum. Mr. George Coffey, m.r.i.a., of the Museum 

XVI 11 


of the Royal Irish Academy, has also from time to time 
been good enough to assist me in various ways. The 
photographs of the cast of the Nigg cross were taken 
by Messrs. M. and T. Scott, of Edinburgh, for Mr. 
D. J. Vallance, the curator of the Museum of Science 
and Art at Edinburgh. 

For the use of electrotypes of blocks I have to give my 
best thanks to the Society of Antiquaries of London,^ 
the Society of Antiquaries of Scotland, ^ the Royal 
Society of Antiquaries of Ireland,^ the Royal Irish 
Academy,"* the Royal Archaeological Institute,^ the 
Cambrian Archaeological Association,*^ the Somerset- 
shire Archaeological Society,^ and the publishers of 
the Antiquary y^ the Reliquary ^^ and the Illustrated 
ArchcEologist,^^ Plates XXVI., XXIX., XXXV., 
XXXVI. , and XXXVII. are from the series of photo- 
graphs taken by Mr. W. G. Moore, of Upper Sackville 
Street, Dublin, for the Royal Irish Academy. 

^ Blocks on pp. 112, 123, 152, 153, 166, 167, 168. 

154, 169, 183, 276 to 278, 274 to 288. 
150, 221. 

39, 40, 45, 47- 

119, 120, 131, 147. 

27 to 36, 181, 185, 187, 189, 191, 259 to 269, 271 to 

273, 283. 

147, 161. 
10 Plate XXXIV. 





ALL the nations at present inhabiting Europe, with 
the exception of the Turks, the Finns, the Mag- 
yars, and the Basques, speak Aryan languages, 
and are to a large extent of Aryan descent,^ although 
their blood has been mixed from time to time with that 
of the Neolithic non-Aryan aborigines. The Celts, 
therefore, belong to the Aryan group of nations, and 
came from the same cradle of the race in Central Asia 
as did the ancestors of the Greeks, Italians, Teutons, 
Slavs, Armenians, Persians, and the chief peoples of 

It has been the fashion amongst persons holding 
what they suppose to be advanced views to dispute 
the fact that the cradle of the Aryans was in Central 
Asia, but this is neither the time nor the place to discuss 

^ The fallacy that identity of lang-uag-e or of culture necessarily 
implies identity of race must be carefully guarded ag-ainst. 


the question. It is sufficient for our purpose to note 
that the successive waves of Aryan conquest entered 
Europe from the east, and that their general direction 
was towards the west. 


The Celts make their first appearance in history at 
the end of the sixth century B.C., when, however, 
they are referred to not by their name as a people but 
by the name of the country they occupied. Thus, 
Hecatceus of Miletus, writing about 509 B.C., mentions 
Marseilles as being a Ligurian city near the Celtic 

Herodotus, writing half a century later, is the first 
historian who uses the word K€\T(k (Celt) as distin- 
guished from KeXriK}'/ (the Celtic region). In the two 
passages' in which the Celts are mentioned, Herodotus 
says that they inhabited the part of Europe where the 
Danube has its source, and that the only other people 
to the westward were the Cynetes or Dog-Men. Both 
Herodotus and Aristotle erroneously supposed that the 
source of the Danube was situated in the Pyrenees. 

Aristotle^ describes the country of the Celts as being 
so cold that the ass is unable to reproduce his species 

Plato, who lived sixty years after the time of Herod- 
otus, classes the Celts with the Scythians, Persians, 
Carthaginians, Iberians, and Thracians, as being war- 
like nations who like wine, and drink it to excess.* 

^ MaaaaXia ttoXls ttjs ALy/xaTLKijs Kara t7]v KeKriK-qv (C. and T, Muellerus, 
Fragmenta Historicorum Grcecorum, Paris, 1841, vol. i., p, 2, fragrn. 22). 
- Bk. ii., chap, xxxiii. ; and Bk. iv., chap, xlix 
^ De Generatione Animaliuni. 
^ Dc Legihiis. 


Pytheas (circa B.C. 300) is the first author who includes 
the part of Europe which was afterwards the Gaul of 
Cassar within the Celtic territory.^ 

According to the earlier historians, the parts of 
Europe occupied by the Celts at the end of the 
fourth century B.C. were the coast of the Adriatic 
from Rimini to Venice, Istria and the neighbourhood 
of the Ionian Gulf, and the left bank of the Rhone 
from the Lake of Geneva to the source of the Danube.'- 

Polybius (B.C. 205-123) gives more definite and satis- 
factory information about the Celts than the somewhat 
vague references made to them by previous writers. 
From him we learn ^ 

(i) That the Celts of upper Italy did not come from the 
Gaul of Caesar, but from the valley of the Danube, and more 
particularly from the countries which border upon the northern 
slopes of the Julian Alps of Noricum. 

(2) That these peoples were primarily divided into Cis- 
alpine Celts and Transalpine Celts, that is to say, into the 
Celts of the Alps and of the north of the Alps. In the third 
century B.C. these latter were already called, more particularly 
by Polybius, by the name of Galati. 

(3) That the Cisalpine Celts, who from a remote period 
long before the fourth century B.C. inhabited the wide plains 
of Lombardy from the Alps to the river P6, were, for the 
most part, an agricultural and sedentary race living in luxury 
and in a state of civilisation without any doubt greatly 
superior to that which could have existed in Gaul at that 

(4) That the Galati, on the contrary, the Transalpine 
Celts, although kinsmen to the former mountaineers, still 
half nomads, shepherds and warriors chiefly, always ready 
to run the risk of a raid, armed from the fourth century with 

^ C. Elton's Origins of English History, p. 25. 
^ A. Bertrand and S. Reinach's Les Celfes, p. 19. 
^ Ibid,, p, 27. 


an iron sword, an iron-headed spear and a shield, Hved under 
the regime of a sort of mihtary aristocracy, as proud as they 
were poor, such as the inhabitants of the Caucasus were not 
half a century ag*o. 

The people who were called Celtce by the earlier 
historians, and Ga/aUu by the more recent writers, 
were also known to the Romans as Galli ; but these 
three separate appellations do not seem to indicate 
any difference of race, and indeed they all have the 
same meaning, viz. a warrior. The Gauls of Caesar's 
time preferred to call themselves by the name which 
he wrote, Celtce.^ 

All the Classical authorities are agreed as to the 
physical characteristics of the Celts with whom they 
were acquainted. The Celts are invariably described 
as being tall, muscular men, with a fair skin, blue eyes, 
and blonde hair tending towards red."^ Such were the 
Gauls who conquered the Etruscans of northern Italy 
in B.C. 396, took Rome under Brennus six years later, 
sacked Delphi in B.C. 279, and gave their name to 
Galatia in Asia Minor. 

It may well be asked what has become of the tall, 
fair-haired Celts who in the fourth century B.C. were 
the terror of Europe? The answer seems to be that 
being numerically inferior to the races which they 
conquered, but did not exterminate, they after a time 
became absorbed by the small, dark Iberians, who 
were the aborigines of France and Spain in the 
later Stone Age. In Great Britain the once warlike 
Celt at last became so effete that he fell an easy 
prey to the Picts, the Scots, the Angles, and the 

^ Prof. J. Rhys' Celtic Britain^ p. 2. • 

■^ C. Elton's Origins of English History, p. 113. 



The physical type of the Celt in Classical Sculpture 
was fixed by the artists of Pergamos, who were com- 
missioned to perpetuate the victories of Attalus I. (b.c. 
241-197) and Eumenes II. (b.c. 197-159) over the Gala- 
tians of Asia Minor.^ The originals of the statues 
executed at this period to decorate the acropolis at 
Pergamon and at Athens have since been popularised 
by means of numerous copies. The statue most familiar 
to everyone is that wrongly called the Dying Gladiator,'' 
but which is really a Gaulish warrior mortally wounded, 
as may be seen by the twisted torque round his neck, 
and the shape of his shield and trumpet. The other 
statues of the same class are the group formerly known 
as Arria and Paetus^ (representing a Gaul committing 
suicide after having killed his wife) and the figures of 
an old man with a young man dead* and a young man 
wounded^ from the defeat of the Gauls by Attalus. 

In all these works of art the Gaulish type is the 
same, the men being tall and muscular, with abundant 
unkempt locks, and an energetic, almost brutal, 
physiognomy, the very opposite of the intellectual 
beauty of the ideal Greek. The type thus fixed by 
eminent artists was handed down from generation to 
generation, until the last years of the Roman empire. 
It may be recognised on the Triumphal Arch at Orange^ 

1 Les Celtes, p. 37 ; H. B. Walters' Greek Art, p. 91 ; and Dr. A. S. 
Murray's History of Greek Sculpture, vol. ii., p. 376. 

'^ In the Museum of the Capitol at Rome; cast in the South Kensington 

'■^ Prof. Ernest Gardner's Handbook of Greek Sculpture, pt. ii., p. 456 ; 
and A. Baumeister's Denkmdler, p. 1237. 

^ At Venice. ° In the Louvre. 

^ A. de Caumont's AhdcMaire d' Archdologie (Ere Gallo-Romaine), 
Second edition, p. 194. 


(Vaucluse), in the south of France, and at the sarco- 
phagus of Ammendola^ in the museum of the Capitol 
at Rome, both of which have derived their inspiration 
from the works of art of the time of the kings of 
Pergamon. Latterly the Gaulish type became that of 
barbarians generally.'- 


From an archaeological point of view the Celtic 
civilisation which existed in Central Europe, certainly 
as far back as 400 B.C., and very probably three or four 
centuries earlier, was that of the Iron Age. The 
Continental antiquaries divide the Iron Age in this 
part of Europe into two periods marked by differences 
in culture. The culture of the Early Iron Age is pre- 
historic, and is called that of '' Hallstatt," after the 
great Alpine cemetery near Salzburg in Austria. 

The culture of the Later Iron Age comes after the 
time when the Celts first make their appearance in 
history, and is known to Swiss and German archae- 
ologists as that of "La Tene," from the Gaulish 
Oppidum at the north end of the Lake of Neuchatel 
in Switzerland. The La Tene culture in the form it 
occurs in France is called " Marnian," and corresponds 
with the " Late-Celtic " culture of Great Britain. 

Hallstatt, from which the Celtic civilisation of the 
earlier Iron Age takes its name, is situated thirty miles 
S.E. of Salzburg in Austria, amongst the mountains 
forming the southern boundary of the valley of the 
Danube. It was a place of great commercial importance 
in ancient times, in consequence of the salt mines in 
the neighbourhood, and because it lay on the great 

^ S. Reinach's Lcs Gaulois da}is I'art antique. 
" S. Reinach's Les Ce//es, p. 38. 


trade route by which amber was brought from the 
mouth of the Elbe to Hatria, at the head of the 

The pre-Roman necropolis of Hallstatt was discovered 
in 1846, and excavations have been going on there at 
intervals ever since. In 1864 M. de Sacken, curator 
of the collection of antiquities in the Vienna Museum, 
published a monograph on the subject, which still 
remains the best book of reference. M. de Sacken 
did not superintend the excavations personally, that 
task having fallen to the lot of George Ramsauer. 
Copies in MS. of Ramsauer's notes on the contents 
of the tombs, and sketches of the antiquities discovered 
in them exist in the Ashmolean Museum at Oxford, 
and in the national museums at Saint-Germain and at 

The Hallstatt finds show very clearly the transition 
from the Bronze to the Iron Age in Central Europe. 

M. Salomon Reinach thus summarises, in his Les 
Celtes dans les Vallees dit Po et du Danithe (p. 129), the 
conclusions arrived at by M. de Sacken : — 

(i) Two distinct races have been buried at Hallstatt ; one 
of which cremated the bodies and the other which practised 
inhumation ; the former showing themselves to have been 
much richer than the latter. 

(2) The people, as represented by their grave-goods, must 
have supported themselves, besides working the salt mines 
(their chief industry), by breeding cattle. The number of 
bones and teeth of animals found in the tombs show that 
they possessed herds. Their agricultural pursuits are proved 
by the presence of numerous scythes and sickles in the 
graves. Slag and moulds from founderies indicate that they 
were metallurgists. 

^ C. Elton's Origins of English History, pp. 46 and 62, and Prof. W, 
Boyd Dawkins' Early Man in Britain^ pp. 417, 466, and 473. 


(3) Among^st the individuals who had been burnt 
the greater part of the men and women displayed a 
relative luxuriousness of toilet appliances, a luxuri- 
ousness which was ministered to by foreign com- 
merce supplying amber from the Baltic, Phoenician 
glass, ivory, embroidery in gold thread and stamped 
gold-leaf of oriental workmanship, used in the 
decoration of the sword-hilts and scabbards. 

(4) On the bronze vessels, side by side with 
the old geometrical ornament, common to them 
and to the Cisalpine vases, are to be seen new 
combinations of symbolical designs which recur 
on the Celtic coinag-e of Gaul. 

Amongst the objects most characteristic of 
the Hallstatt culture are : — 

(i) Daggers, or short swords, with a pointed 
blade of iron and a bronze handle having two 
little projections at the top terminating in round 
knobs and resembling the antennae of an insect. 

(2) Long double-edged swords with an iron 
blade made in imitation of the leaf-shaped swords 
of the Bronze Age, having the edges slightly 
curved outwards in the middle, but not having so 
sharp a point as the Bronze Age sword, and 
being much longer. The hilts have a massive 
pommel encrusted with ivory and amber, and 
ornamented with gold-leaf. 

(3) Pails, or situlse, of thin bronze plates orna- 
mented with figure subjects executed in repousse 
work, and exhibiting a peculiar style of art which 
Dr. Arthur Evans thinks the Celts borrowed from 
the Veneti, the ancient Illyrian inhabitants of the 
north of the Adriatic, who, in their turn, had 
come under Hellenic influence whilst the amber 
trade route between Greece and the Baltic passed 
through Hatria. 





J> d^ 






Dr. Arthur Evans^ divides the Hallstatt remains into 
an earlier and a later group, the former dating from 
about 750 to 550 B.C. During the later period, from 
550 B.C., he thinks there was a tendency for the 
typically Gaulish or Late-Celtic culture to overlap that 
of the Early Iron Age. The Gallo-Italian tomb of a 
Celtic chieftain, found in 1867 at Sesto-Calende," at the 
south end of Lago Maggiore, illustrates the transition 
from the Hallstatt to the La Tene culture. Amongst 
the grave-goods were a situla with figure subjects in 
repousse metalwork and a short pointed iron sword 
having a handle furnished with antennae, like those 
from Hallstatt. 

In addition, there were the remains of a chariot, 
horse -trappings, a bronze war- trumpet, helmet and 
greaves, and iron lancehead, such as we should expect 
to find buried with a Celtic warrior in the Iron Age in 
Gaul or Britain. 

La Tene (which gives its name to the modified and 
later form of Hallstatt culture as it existed in Central 
Europe from about 400 B.C., when the name of the 
Gaul superseded that of the Celt, to the time of 
Caesar's conquest) is a military stronghold, or oppidum, 
situated at the N.E. end of the Lake of Neuchatel, 
commanding an important pass between the upper 
Rhone and the Rhine. The remains at La Tene 
were first explored by Colonel Schwab in 1858, and 
subsequently by E. Vouga in 1880. The objects 
derived from this remarkable site are to be seen in 
the public museums at Bienne, Neuchatel, and Berne ; 
and in the private collections of Colonel Schwab, 

^ Rhine! Lectures on the "Orig-ins of Celtic Art," Lectvire II., as 
reported in the Scotsman for December 12th, 1895. 
'■^ S. Reinach's Les Celtes, p. 49. 


Professor Desor, E. Vouga, Dardel Thorens, and 
Dr. Gross.^ 

According to Dr. Arthur Evans, the date of the 
culminating epoch of Gaulish civilisation, as repre- 
sented by the antiquities from La Tene, is probably 
the third century B.C. It was at this period that the 
earlier foreign elements derived from Hallstatt, and 
even from countries further afield, became thoroughly 
assimilated, and the style of art called Late-Celtic 
began to take definite shape. 

The typical arms found at La Tene are : — 

(i) A long sword with a double-edged iron blade having 
a blunt point. The length and flexibility of the blade made 
it useless for thrusting in the way which was possible with 
the shorter and more rigid leaf-shaped sword of the Bronze 
Age, so the pointed end was abandoned. 

(2) Lances with an iron point often of a peculiar curved 

(3) An oval shield of thin bronze plates ornamented with 

(4) A horned helmet of bronze. 

The characteristic La Tene ornament is found chiefly 
on the sword-sheaths, the helmets, and the shields. 
The La Tene fibulce are derivatives of the ''safety-pin," 
and usually have the tail-end bent backwards, as in 
the Marnian fibulae in France and the Late-Celtic 
fibula in England. 

The Gaulish culture in France corresponding with 
that of La Tene in Switzerland has been called 

^ The remains are fully described in Dr. F. Keller's Lake Divelliiigs ; 
Dr. R. Munro's Lake Dwellings of Europe ; E. Vouga's Les Helvetes 
a la Tene ; and Dr. Gross' La Tene un Oppiduin Helvete. 

'•^ With flame-like undulating- edges "so as to break the flesh all in 
pieces" (C. Elton's Origins of English History, \). i t6). 





'^ Marnian " by the French archseologists because the 
principal remains of this period have been found in 
the cemeteries of the Department of the Marne. A 
list of seventy-two such Marnian cemeteries (some of 
which contained as many as 450 graves) is given by 
A. Bertrand in his A^xheologie Celtiqite et Gaiiloise^ 
p. 358. The objects obtained from these cemeteries 
are fully illustrated in the Dictionnaive Archeologique 
de la GaidCy and in Leon Morel's La Champagne 

Ornament on Bronze Sword-sheath from La Tene 

soiiterraine {Album). The best collections are those in 
the Museum of Saint-Germain and the British Museum. 
M. Bertrand fixes the date of the Marnian cemeteries 
at from 350 to 200 B.C., the period between the time 
when bronze weapons ceased to be used, and the intro- 
duction of a national coinage into Gaul. 

From the point of view of art, two of the most in- 
teresting burials discovered in the Departement du 
Marne are those at Berru^ and Gorge-Meillef' of 

^ A. Bertrand's Arch^ologie Celtique et Gauloise, p. 356. 

^ E. Fourdrig-uier's Double St'pulture Gauloise de la Gorge-Meillet. 



warriors interred with their chariots, horses, and com- 
plete military equipment, including two bronze helmets, 
which show the kind of decoration prevalent at the 
period, and afford a link between the Marnian style 
in Gaul and the Late-Celtic style in Britain. The 

Gaulish Helmet of Bronze from Gor^e-Meillet 

burials at Berru and Gorge-Meillet correspond very 
nearly with those at Arras, Danes' Graves, and else- 
where, in the portion of Yorkshire occupied by the 
Celtic tribe of the Parisi. 

The Marnian cemeteries belong to the second Iron 
Age of Central Europe after 400 B.C., but in the com- 
mune of Magny Lambert (Cote-d'or), near the source 

CASQUE DE BERRU. i mar ke. 1 decouvert dans ur; Cimetiere Gadoi; 


/ ^ 5 / 3 


of the river Seine, tumuli have been opened containing 
long iron swords and bronze situlas of distinctly Hall- 
statt type. 

Dr. Arthur Evans thinks that the older, or Hallstatt, 
culture of Central Europe was gradually modified and 
transformed into the LaTene, Marnian, and Late-Celtic 
stages of culture, in consequence of the foreign in- 
fluence exercised by the continual flow of Greek 
commerce into eastern Gaul from the sixth century B.C. 
onwards. Ample evidence of this commercial inter- 
course is afforded by the discovery of tripods, hydrias, 
oenochces, and painted vases of Greek workmanship 
associated with Gaulish burials,^ as at Gr^ekwyl, near 
Berne in Switzerland, at Somme-Bionne (Marne), at 
Rodenbach in Bavaria, and at Courcelles-et-Montagne 

The great difficulty in understanding the evolution of 
Celtic art lies in the fact that although the Celts never 
seem to have invented any new ideas, they professed 
an extraordinary aptitude for picking up ideas from the 
different peoples with whom war or commerce brought 
them into contact. And once the Celt had borrowed 
an idea from his neighbour, he was able to give it such 
a strong Celtic tinge that it soon became something so 
different from what it was originally as to be almost 

Polybius gives the following picture of the Cisalpine 
Gauls : — 

*' These people camp out in villages without walls, and 
are absolutely ignorant of the thousand things that make 
life worth living. Knowing no other bed than straw, only 
eating flesh, they live in a half-wild state. Strangers to 

^ A. Bertrand's Archdologie Celtique et Gauloise, pp. 328 to 347 ; see 
also L. Lindenschmit's Die Alterthihner xinserer heidnischen Vorzeit, 
Mainz, 1858, etc. 


everything which is not connected with war or agricul- 
tural labour, they possess neither art nor science of any 

The tendency of the Celt to copy rather than invent 
is brought out most clearly in their coinage. M. A, 
Bertrand^ says : — 

" Were they settled in Macedonia they imitated with more 
or less success the tetradrachms of Philip and of Audoleon, 
king of Paeonia ; did they advance towards Thrace, they 
copied the tetradrachms of Thasos. The Senones of 
Rimini took for their model the Roman and Italian aes grave ; 
in the north of Italy, finding themselves in contact with 
nations who used the monetary system of the drachm and 
its multiples and divisions, the Gauls copied them until 
the time they were driven back on the Danube. In Liguria 
they copied the drachms of Massalia. Were they encamped 
on the banks of the Danube in Noricum, or in Rhaetia, they 
again copied the monetary systems of their neighbours. 
The tetradrachms of the Boii on which are inscribed the 
name of 'biatec,' one of their chiefs, reproduced the type 
of the last Roman of the family of Fufia struck between the 
years 62 and 59 B.C. In a word, the same habit of imitation 
is found everywhere in the cradle of Gaulish numismatics 
properly so-called ; on the left bank of the Rhine, it was the 
gold staters of Philip which served as the model for gold 
pieces and sometimes for silver ; in Aquitaine, it was the 
coins of Emporia, Rhoda, and Massalia. Armorica and the 
frontier countries were the first who adopted for their coinage 
types which can be called national, although still reflecting 
those imitated from the Macedonian staters. Let it be 
noticed that we are in one of the most Celtic parts of Gaul : 
it is therefore natural that the difference in genius between 
the two races of Celts and Gauls should manifest itself most 

^ Archdologie Celtique et Gauloise, p. 387. 



The aborigines of Europe, who were driven west- 
ward by the successive waves of Aryan conquest, 
appear to have been in the Neolithic stage of culture, 
and they are identified by Prof. W. Boyd Dawkins 
with the Iberians mentioned by Strabo. Prof. W. 
Boyd Dawkins gives a map in his Early Man in 
Britain (p. 318) showing the relative distribution of the 
Iberic, Celtic, and Belgic races in the historic period. 
In this map the Iberians occupy the north of Africa, 
the west of Spain and France, the country round 
Marseilles, the whole of Wales, and the south-west 
of Ireland. The Celts follow behind to the eastward, 
pressing the Iberians towards the Atlantic. 

In the opening address of the Antiquarian Section 
at the meeting of the Royal Archceological Institute at 
Scarborough in 1895, Prof. Dawkins said : — ^ 

"The theory that the Neolithic inhabitants of the British 
Isles are represented by the Basques and small, dark Iberic 
population of Europe generally, has stood the test of twenty- 
five years' criticism, and still holds the field. From the side of 
philology it is supported by the fact pointed out by Inchauspe 
that the Basque word aitz for stone is the root from which 
the present names of pick, knife, and scissors made of iron 
are derived. This of itself shows that the ancestors of the 
Basques were in the Neolithic stage of culture. The name 
of Ireland, according to Rhys," is derived from Iber-land 
(Hibernia), the land of the Iberians, or sons of Iber. The 
evidence seems to be clear: i. That the Iberians were the 
original inhabitants of France and Spain in the Neolithic 
age, and the only inhabitants of the British Isles ; 2. That 

^ ArchcEoIogical Journal^ vol. lii. , p. 342. 
^ Celtic Britain, p. 262. 


they were driven out of the south-eastern parts of France 
and Spain in the NeoHthic age; (3) That they are now 
amply represented by the small dark peoples in the Iberian 
Peninsula, and in the island which bears their name, and in 
various other places in Western Europe, where they con- 
stitute, as Broca happily phrases it, 'ethnological islands.' 
The small, dark, long-headed Yorkshiremen form one of 
these islands." 

Let us pause for a moment to consider the stage of 
culture attained by the Neolithic aborigines of Britain 
whom the Celts found here on their first arrival. The 
houses in which Neolithic man lived are of two kinds : 
(i) pit dwellings dug to a depth of from seven to ten 
feet deep in the chalk, like those at Highfield,^ near 
Salisbury, explored by Mr. Adlam in 1866; and (2) hut 
circles like those at Cam Bre near Camborne," in 
Cornwall, excavated by Mr. Thurstan C. Peter, and on 
Dartmoor," excavated by the Rev. S. Baring-Gould and 
Mr. R. Burnard. In many cases the villages are forti- 
fied by a wall of rubble stone, as at Grimspound, on 
Dartmoor. Neolithic dwellings have also been explored 
by Mr. George Clinch, at Keston, in Kent.^ 

Neolithic man supported himself by the chase and 
by fishing, and also was a farmer in a small way, 
growing wheat and cultivating flax. He had domesti- 
cated the sheep, goat, ox, hog, and dog. He could 
spin, weave, mine flint, chip and polish stone im- 
plements and make rude pottery. 

^ W. Boyd Dawklns' Early Man in Britain,'^. 267; and E. T. Stevens' 
Flint Chips, p. 57. 

■^ R. Burnard in Trans, of Ply month Inst., 1895-6; and T. C. Peter 
\x\ Jour. R. Inst, of Cornwall., No. 42. 

•' Reports of Dartmoor Exploration Committee in the Trans, of 
Devonshire Assoc, for Advancement of Science. 

^ Proc. Soc. Ant. land., ser 2, vol. xii., p. 258, and vol, xvii., p. 216. 


He buried his dead in long barrows, chambered 
cairns, and dolmens. Cremation was not practised, 
and it was usual to inter a large number of bodies 
in a chamber constructed of huge stones. 

Such was the aboriginal inhabitant with whom the 
first Celtic invader had to contend. I S3.y first Celtic 
invader advisedly, for there was a second Celtic in- 
vasion at a later period. The vanguard of the Celtic 
conquerors are called by Prof. J. Rhys, in his Celtic 
Britain (p. 3), ^'Goidels," to distinguish them from 
the ''Brythons," who constituted the second set of in- 
vaders. The modern representatives of the Goidels 
are the Gaelic-speaking population of the Highlands 
of Scotland, Ireland, and the Isle of Man ; whilst the 
descendants of the Brythons now inhabit Wales, 
Cornwall, and Brittany. At the time of the Roman 
occupation the Brythonic tribes inhabited the whole 
of England with the exception of the districts now 
occupied by Cumberland, Westmoreland, Devon, and 
Cornwall. The most important of these Brythonic 
tribes were the Brigantes and Parisi of Yorkshire, the 
Catuvelauni of the Midland Counties, the Eceni of 
the eastern counties, the Attrebates of the Thames 
Valley, and the Belg^, Regni, and Caution in the south. 
The south of Scotland was in possession of the 
Dumnoni and Otadini, who were Brythons, as were 
also the Ordovices of Central Wales. The Ivernians 
still held their own in the north of Scotland. The 
remainder of Great Britain was inhabited either by 
pure Goidels or by Goidels who had mixed their blood 
with the Ivernian aborigines. 

As Prof. J. Rhys has pointed out in his Celtic 
Britain (p. 211), the soundest distinction between the 
Goidels and the Brythons rests on a peculiarity of 


pronunciation in their respective languages. In the 
corresponding words in each language where the Bry- 
thons use the letter P, the Goidels use Qv. Hence 
they have been termed the "P and Q Celts." The 
most familiar instance of this is where the Welsh use 
the word ap to mean son of^ and the Gaels use viae. 
The older form of mac found on the Ogam-inscribed 
monuments of Ireland, Scotland, Wales, and the West 
of England is maqvi^ as in the bi-literal and bi-lingual 
inscribed stone at St. Dogmael s, in Pembrokeshire, 
where the Latin ''sagrani fili cvnotami " has as an 
equivalent in Ogams " sagramni maqvi cynatami." 
In modern Welsh map^ or mah^ has been shortened by 
dropping the 7;?, and in Gaelic the v of maqvi has been 
dropped, and the q made into c. 

So much for the philological differences between the 
Goidel and the Brython. They can also be distinguished 
archasologically, the former as being in the Bronze Age 
stage of culture, and the latter in the Early Iron Age 
when he arrived in Britain. In a subsequent chapter 
we shall have to deal with the Brythonic Celt, but at 
present we are concerned exclusively with the Goidel. 

The Neolithic inhabitants of this country, whom the 
Goidelic Celts found here on their arrival, were ethno- 
logically a small dark-haired, black-eyed race, with 
long skulls of a type which is still to be seen amongst 
the Silurians of South Wales. ^ The ethnological 
characteristics of the Goidels were entirely different : 
they were tall, fair-haired, round-headed, with high 
cheek-bones, a large mouth, and aquiline nose. In 
studying the past much must necessarily be more or 
less conjectural, and we can never hope to see otherwise 
than "as in a glass darkly." As far, however, as it is 

^ Boyd Dawkins' Early Man in Britain, chapter ix. 


possible to ascertain the facts, it appears probable that 
the advancing wave of Goidelic Celts did not entirely 
overwhelm the aborigines or drive them before it. 
Most likely the big Goidels made the small Iberians 
^'hewers of wood and drawers of water," and in time 
either absorbed them or themselves became absorbed. 


Actual dates in years can only be ascertained by 
means of historical documents, and therefore no 
chronology of the ages of Stone, Bronze, and Iron 
is possible except where contact can be established 
between the prehistoric (or non-historic) races living 
in those stages of culture in Northern and Central 
Europe, and the more advanced civilisations on the 
shores of the Mediterranean and in Asia. Long 
before direct contact took place between the northern 
barbarians and the Egyptians, Assyrians, Phoenicians, 
Greeks, Romans, and other great nations of antiquity, 
through invasions or immigrations, a more indirect 
contact must have existed for centuries, owing to the 
trade in such things as amber, gold, bronze, and tin. 
Dates have been fixed approximately by the finding 
of imported objects in different countries, and by 
studying their geographical distribution. Other almost 
untouched fields of investigation which would help to 
solve many of the problems of prehistoric chronology, 
are the migration of symbols and patterns and com- 
parative ornament. 

The attempts that have been made to fix the duration 
of the Ages of Stone and Bronze in actual years are 
at the best mere guesses, but it may be worth while 
stating the conclusions arrived at by some of the leading 
European archaeologists, so as to give a rough idea of 


the time at which bronze was in use for the manufacture 
of implements and weapons in different countries. 

Egypt during the greater part of its existence as a 
civilised nation was in the Bronze Age. The copper 
mines of the Sinaitic peninsula were worked as early 
as the Fourth Dynasty, as is proved by the rock in- 
scriptions of Sneferu (b.c. 3998-3969) at Wady Mag- 
hera.^ Bronze was certainly used by the ancient 
Egyptians in the fourteenth century B.C., and in the 
tomb of Queen Aah Hotep, although bronze weapons 
were found, iron was conspicuous by its absence, indi- 
cating that the latter metal had not come into general 
use in the fifteenth century B.C. 

The Mycenaean civilisation in the ^gean was of 
the Bronze Age, and Prof. Flinders Petrie places its 
flourishing period at about 1400 B.c.^ Bronze continued 
in use in Greece until the time of the Dorian invasion, 
B.C. 800. 

In dealing with the local centres of the bronze in- 
dustry. Prof. Boyd Dawkins^ recognises three distinct 
local centres in Europe. 

(i) The Uralian, or Eastern — Russia. 

(2) The Danubian, or Northern and Central — Scandinavia, 

(3) The Mediterranean, or Southern — Greece, Italy, France, 

Dr. Oscar Montelius^ gives the following tentative 
dates for the duration of the Bronze Age in these 
areas : — 

^ Petrie's Hist, of Egypt, vol. i., p. 31. Article on "The Age of 
Bronze in Egypt," in L' Anthropologie for January, 1890, translated in 
the Smithsonia7i Report for 1890, p. 499. 

'^ Jour, of Hellenic Studies, vol. xii. , p. 203. 

2 Early Man in Britain, p. 414. 

* Mat^rianx pour V histoire primitive de Vhoynme, pp. 108-113. 


The Caucasus. — The Massagete were, according to Herod- 
otus, still using bronze in the sixth century B.C. 

Greece. — Bronze Age civilisation of Mycenae, 1400 to 
1000 B.C. 

Italy. — Terramare of Bronze Age, twelfth century B.C. 
Iron introduced in ninth or eighth century B.C. 

Sca7idi7iavia a?id Germany. — Bronze Age begun in fifteenth 
century B.C., and ended in fifth century b.c. 

Worsaae^ places the beginning of the Bronze Age in 
Scandinavia five centuries later than Montelius, i.e. 
1000 B.C. 

Dr. Naue- dates the Bronze Age in Upper Bavaria 
from 1400 B.C. to 900 B.C. 

As regards Great Britain, there is no reason for 
supposing that the Brythonic Celts of the Early Iron 
Age arrived in this country much before B.C. 300, 
which date would terminate the Bronze Age, at all 
events in southern England. The date of the begin- 
ning of the Bronze Age in Britain can only be 
surmised. If, as we hope to be able to prove, much of 
the art of that period here can be traced to a Mycenaean 
origin there is no reason why the Bronze Age in Britain 
should not have commenced shortly after the spiral- 
motive patterns were transferred from ancient Egypt 
to the ^gean, say, about 1400 B.C., and thence to 
Hungary, Scandinavia, and other parts of Europe. 
It is not impossible, nay, it is even probable, that the 
Bronze Age may have lasted a thousand years in 
Britain, beginning B.C. 1300, and ending B.C. 300. 

^ The Industrial Arts of Denmark^ p. 41. 

^ Dr. Arthur Evans' review of Dr. Julius Naue's Die Bronzezeit iti 
Ohayer?i in the Academy for April 27th, 1895. 



AS we have already observed, the Goidelic Celts 
were in the Bronze Age stage of culture when 
they landed in Britain. Let us now inquire into 
the nature of the materials available for the study of 
the Pagan Celtic art in the Bronze Age. 

The remains of this period may be classified, accord- 
ing to the nature of the finds, as follows : — 

(i) Sepulchral remains. 

(2) Remains on inhabited and fortified sites. 

(3) Merchants' and founders' hoards. 

(4) Personal hoards, that is to say, finds of objects pur- 
posely concealed, either in times of danger, or buried as ex 
voto deposits. 

(5) Finds of objects accidentally lost. 

(6) Sculptured rocks and stones. 

The art of the Bronze Age in Europe is both of a 
symbolical and decorative character. The principal 
symbols employed are : — 

The Swastika. 

The Triskele. 

The Cup and Ring. 

The Ship. 
The Axe. 
The Wheel. 




It is probable that most of these were connected with 

The chief decorative art motives which were pre- 
valent during the Bronze Age are as follows : — 

The Chevron. 

The Concentric Circle. 

The Spiral. 

The Winding Band. 

With the introduction of bronze into Britain an 
entire change took place in the burial customs of the 
people. The long barrows with their megalithic 
chambers and entrance passages gave place to round 
barrows containing cists constructed of comparatively 
small slabs of stone, and having no approach from the 

Although burial by inhumation still continued to be 
practised, cremation was adopted for the first time. 
The proportions of unburnt to burnt bodies found in 
opening barrows in different parts of England vary 
according to Thurnam*^ thus : — 







The survival of the practice of inhumation to so 
large an extent would seem to indicate that the bronze- 
using Goidels amalgamated with the Neolithic aborig- 
ines rather than exterminated them. 

The unburnt bodies were usually buried in a doubled- 

^ See J. J. A. Worsaae's Danish Arts, p. 68. 
'•^ ArchcBologiay vol. xliii. , p. 310. 






58 .. 





up position, and sometimes an urn was placed near the 
deceased. When the body was cremated the ashes 
were placed in a cinerary urn, and the grave-goods 
most commonly consisted of smaller pottery vessels, a 
bronze dagger or razor, and a stone wrist-guard. 
Occasionally flint implements and polished stone axe- 
hammers have been found with burials of this type, 
but it does not necessarily follow, in consequence, that 
bronze was unknown at the time. 

The sepulchral pottery derived from the round 
barrows of the Bronze Age supplies us with ample 
material for studying the art of the period. 

The principal collections are to be seen in the British 
Museum and the museums at Devizes, Sheffield, Edin- 
burgh, and Dublin. These have been derived from 
the barrows opened by Sir R. Colt Hoare in Wiltshire, 
T. Bateman in Derbyshireand Staffordshire, Rev. Canon 
Greenwell and the Rev. J. C. Atkinson in Yorkshire, C. 
W^arne in Dorsetshire, and W. C. Borlase in Cornwall. 

The pottery from the round barrows exhibits an 
endless variety of form, but as regards their suggested 
use, they may be divided into four classes, namely : — 

(i) Cinerary urns. 

(2) Food-vessels. 

(3) Drinking-cups. 

(4) Incense-cups. 

There is no doubt as to the use for which the cinerary 
urns^ were intended, because they are found filled with 
burnt human bones, sometimes placed in an inverted 
position upon a flat stone, and sometimes mouth up- 
wards. The cinerary urns vary in height from 6 inches 
to 3 feet, and the most common shape resembles that 

^ Greenwell's British Barroii's, p. 66. 




HEIGH r 3^ INS. 







of an ordinary garden flower-pot, with a deep rim round 
the top, probably to give the vessel greater strength. 

The so-called food-vessels^ have received this name 
because they are believed to have contained food for 
the deceased in the next world. In support of this 
theory it may be mentioned that remains of substances 
resembling decayed food have been found in some of 
the vessels in question. Urns of the food-vessel type 
are shaped like a shallow bowl, and they vary in height 
from 3 to 8 inches. They are usually found placed 
beside the deceased. 

The use of the so-called drinking-cups' is suggested 
more by the form, which resembles that of a mug, or 
beaker, slightly contracted in the middle, than by any 
actual facts connected with their discovery. They are 
generally placed near the deceased. The height of 
the drinking-cups varies from 5 to 9 inches. The 
Hon. J. Abercromby, f.s.a. (Scot.), has recently 
published an elaborate monograph on the drinking- 
cups of the Bronze Age entitled "The Oldest Bronze 
Age Ceramic Type in Britain ; its close Analogies on 
the Rhine ; its Probable Origin in Central Europe."^ 

Incense-cups were conjectured by Sir R. Colt Hoare 
and the earlier archaeologists to have been used for 
burning some aromatic substance during the funeral 
rites. The view taken by the late Mr. Albert Way, 
and supported by Canon Greenwell,^ is that they were 
for carrying burning wood to light the funeral pile. 
The incense-cups are the smallest of the sepulchral 
vessels of the Bronze Age, being only from i to 3 inches 
high. The shape is like that of a little cup. The sides 

^ British Barrows, p. 84. '- Ibid., p. 94. 

^ Jour. Anthropolog. Inst, vol. xxxii. , p. 373. 
■* British Barrows, p. 81. 


are sometimes perforated. The incense-cups are often 
found inside the cinerary urns. 

Canon Greenwell states that the urns of the four 
different types were found associated with unburnt and 
burnt bodies in the barrows opened by him on the 
Yorkshire wolds in the following proportions : — 

Unburnt. Burnt. 

Cinerary urns . 12 ... 9 

(of cinerary urn type, (containing 

but without ashes) burnt bones) 

Food-vessels . . 57 ... 16 

Drinking-cups . 22 ... 2 

Incense-cups. . none ... 6 

The geographical distribution of the different types 
of sepulchral urns, as far as at present ascertained, 
is as follows : Food-vessels are most common in 
Yorkshire, and most rare in Wiltshire and the south 
of England generally. Drinking-cups are found all 
over Great Britain,^ and it is the type of urn which 
varies least. Incense -cups are found with greater 
frequency in the south of England than in the north. 

Now as to the decorative features of the sepulchral 
pottery of the Bronze Age in Great Britain. 

The sepulchral urns are made of coarse clay moulded 
by hand — not turned on a lathe — and imperfectly baked 
by means of fire. The decoration was executed whilst 
the clay was moist, either by 

(i) The finger-nail. 

(2) An impressed cord. 

(3) A pointed implement. 

(4) Stamps of wood or bone. 

Besides incised patterns produced by these methods, 
the ornament was sometimes moulded in relief and 

^ See map g-'iven by the Hon. J. Abercromby in iht^Jour. Anthropolog. 
Inst., vol. xxxii. , pi. 24. 





sometimes sunk, and the incense-cups often have orna- 
mental perforations. 

With the exception of the circles found on the 
bottoms of some of the incense-cups the decoration 
consists entirely of straight lines running more often 
diagonally than either horizontally or vertically. The 
same preference for diagonal lines will be observed 
in the key patterns in the Irish MSS. of the Christian 

(a) Party per Chevron. 
{b) Party per Saltire. 
(r) Chevron. 
{d) Saltire. 
(e) Indented. 
(/) Dancettee. 

Fig. I. 

period, and led, as we shall see in a subsequent chapter, 
to those modifications of the Greek fret which are 
characteristically Celtic. 

Of the hundreds and hundreds of sepulchral urns of 
the Bronze Age that have been found in Great Britain 
no two are exactly the same either in size, form, or 
decoration. The fertility of imagination exhibited in 
the production of so many beautiful patterns by com- 
bining diagonal straight lines in every conceivable way 
is really amazing. On examination it will be found 



that, complicated as the patterns appear to be, the 
chevron or zigzag is at the base of the whole of them. 
We use the heraldic terms for the sake of convenience; 
their meaning will be understood by a reference to 
Fig. I. 

It will be seen that the chevron consists of two 
straight lines or narrow bars inclined towards each 
other so as to meet in a point, the form thus produced 
being that of the letter V. Now the chevron, or V, is 
capable of being combined in the following ways : — 

W. — Two chevrons, with the points facing in the same 

direction, placed side by side. 
-Two chevrons, with the points facing in opposite 

directions, placed with the open sides meeting. 
X. — Two chevrons, with the points facing in opposite 

directions, placed with the points meeting. 

By repeating the W, (^, and X, each in a horizontal 
row, the patterns shown on Fig. 2 are obtained. 


l\ AAA/\/VI 






(a) The Triang-Ie, or Chevron 

(/?) The Lozeng-e Border. 
(r) The Saltire Border. 
{d) The Hexag-on Border. 


It will be noticed that the same pattern results from 
repeating a series of A's in a horizontal line as from 
repeating a series of X's, so that in order to distinguish 
the lozenge border from the saltire border, it is necessary 



to introduce a vertical line between each pair of Xs. 
The hexagon border is derived from the lozenge by 
omitting every other X. 

It is a principle in geometrical ornament that for 
each pattern composed of lines there is a corresponding 
pattern in which bars of uniform width are substituted 
for lines. Another way of stating the same propo- 
sition is, that for each pattern composed of geometrical 


{a) Line Chevron Border. 
(b) Bar Chevron Border, 
(r) Surface Pattern, produced 

by repeating- either of 

the preceding-. 

figures (squares or hexagons, for instance) there is a 
corresponding pattern produced by moving the figures 
apart in a symmetrical manner so as to leave an equal 
interspace between them. This principle is illustrated 
by Fig. 3, where a zigzag bar is substituted for the 
zigzag line of the triangle or chevron border. 

Then, again, another set of patterns may be derived 
from those composed of lines or plain bars, by shading 
alternate portions of the design as in chequer-work. 
Thus on Fig. 4 are shown three different ways of 
shading the chevron border, and on Fig. 5 the method 
of shading the patterns on Fig. 3. 



^■ mwvM 



Fig. 4. 

Fig-. 5- 
Y\g, 4.— ((7) Line Chevron Border. 

(6, r, and d) Different Methods of Shading- {a). 
Fig. S-— («) Bar Chevron Border. 

[b] The same as (a), but shaded. 

(r) Surface Pattern, produced by repeating [b). 

A few new patterns (see Fig. 6) may be produced by 
placing the chevron with the point of the V facing to 
the right or left, thus, < or > , instead of upwards or 
downwards, thus, A oi" V. 

■ i>>»»»»»:j 


(<'/) Chevron Border, with V's placed 
thus, > > . 

{b) The same as («), but with a hori- 
zontal line through the points 
of the V's. 

(r) The same as (a), but shaded. 

(d) The same as {b), but shaded. 




Fig. 6. 

Figs. 7 to 10 give the triangular patterns, plain and 
shaded, produced by repeating the chevron border 
(see Fig. 2, a). 










Fig. 7. Fig. 8. 

Fig. 7. — {a) Single Border, composed of Triangles. 

[b) Double Border, composed of Triangles, with the points of 

all the Triangles meeting", 
(r) Surface Pattern, composed of Triangles, with the points of 
all the Triangles meeting. 
Fig. 8. — («, b, and c) The Patterns shown on Fig. 7, shaded. 










w w w w 

^ Www WW 

Fig. 9. Fig. 10. 

Fig. 9. — (a) Double Border, composed or Triangles, with the points or 
the Triangles in one row falling in the centres of the bases 
of Triangles in the row above. 
(6) Surface Pattern, composed of Triangles, arranged in the 
same way as in the preceding. 

Fig. 10. — [a and b) The Patterns shown on Fig. 9, shaded. 



The patterns derived from the lozenge are shown on 

Figs. II to i8. 

" IZ 



iAy\y\y\y\/s i 

F"i_iif. 1 1. 

Fig". 12. 

Fig-. II. — {a) Lozeng-e Border, composed of two sets of Chevrons, with 
their points facing- in opposite directions. 
{b) The same as {a), but with the Chevrons set apart. 
(r) The same as (a), but with bars substituted for lines. 
(d) The same as (6), but with bars substituted for Hnes. 

Fig-. 12. — {a) Lozeng-e Border, with Triang-les or Chevrons, shaded. 
(b) Lozeng-e Border, with Lozeng-es shaded, 
(r) The same as Fig-, ii (r), but shaded. 
(d) The same as Pig", ii (d), but shaded. 








Fig-. 13. 

Fig. 13.— (a) Surface Pattern, produced by repeating- the Bar-Chevron 
Border, so that the points of all the Chevrons meet. 
{b) The same as (a), but with the Chevrons set apart. 

Fig. 14. 

Fig. 14. — {a) The same as Fig, 13 («), but shaded. 
[b) The same as Fig. 13 (a), but shaded. 





FiR- 15- 

Fig-. 1 6. 

Fig. 15. — [a) Line Lattice-work Surface Pattern, produced by the repe- 
tition of either the Chevron Border, Fig-. 2 («), or the 
Lozenge Border, Fig. 2 {b). 

{b) The same as («), but shaded. 

Fig-. 16. — {a) The same as Fig-. 15 (6), but with shaded Lozeng-es of two 
different sizes. 
[b) Lattice- work Surface Pattern ; the same as Fig. 15 (b), but 
with diagonal white bars instead of Unes. 









Fig. 17. Fig. 18. 

Fig. 17. — (a) Bar Lattice-work-Surface Pattern; the same as Fig. 15 (a), 
but with diagonal bars instead of lines. 
(b) The same as (a), but shaded. 
Fig. 18. — (a) Surface Pattern, produced by repeating Fig. 11 {c). 
{b) The same as (a), but shaded. 

The patterns derived from the saltire are shown on 
Fig. 19. 






Fig. 19. 

(«) Saltire Border Pattern. 

{b, c, d) Saltire Border Pattern, 

shaded in different ways. 
{e) The same as (a), but with 

bars instead of lines. 



The patterns derived from the hexagon are shown on 
Figs. 20 and 21. 


!^ .^^^ r<yi^ „ ^^ 

Fig-. 20. 

Fig. 21. 

Fig. 20. — {a) Hexagon Border Pattern, derived from the Lozenge Border, 

Fig. 2 (b), by leaving out every other X. 

(6) The same as («), but with the Triangles shaded. 

(c) The same as {a), but with the Hexagons shaded. 

(d) Surface Pattern, composed of Hexagons and Triangles ; 

produced by repeating (c), so that the Hexagons in one 
horizontal row adjoin the Triangles in the next. 

Fig. 21.— («) Hexagon Surface Pattern, probably derived from Fig. ii (6), 
by drawing straight lines between the points of each of 
the Chevrons. 
{b) The same as {a), but with bars instead of lines, and having 
the Hexagons shaded. 

The variations in the practical appHcation of the 
chevron patterns, which have been described above, 


to the decoration of the sepulchral pottery of the 
Bronze Age, are produced in the following ways : — 

(i) By placing the chevrons {a) horizontally, or {b) verti- 

(2) By making the chevrons of different sizes. 

(3) By altering the angle of the chevrons, i.e. making the 

points more acute or more obtuse. 

(4) By shading some parts of the pattern whilst other parts 

are left plain. 

(5) By using different methods of shading, such as plain 

hatching, cross-hatching, dotting, etc. 

(6) By combining the chevrons with horizontal and vertical 


(7) By arranging the patterns in horizontal bands of different 


In a few cases^ hexagonal figures occur in the 
decoration of the urns, but the patterns do not belong 
to the true hexagonal system of ornament. The hexagons 
were arrived at by leaving a space between the triangles 
of the chevrons, as on a drinking-cup found at Rhos- 
heirio,^ Anglesey. 

The decoration of the urns is generally confined to 
the exterior, the only exceptions being the interiors of 
the lips of some of the examples and the crosses in 
relief found on the bottoms inside of cinerary urns from 
Wilts, Dorset, and Sussex. 

The incense-cups have occasionally ornament on the 
bottoms of them which, like the crosses just mentioned, 
may have a symbolical significance. 

Some of the urns from Ireland, Scotland, and the 
Isle of Man, are very beautifully decorated with sunk 
triangles and ovals. 

^ Ll. Jewitt's Grave Mounds and their Contents^ p. io8. Folkton, 

- ArchcBologia Cambrensis, 3rd sen, vol. xiv. , p. 271; British BarroTvs, 
p. 70. 


The different types of urns are not all equally highly 
ornamented. The large flower-pot-shaped cinerary 
urns have least decoration, being sometimes quite 
plain, but in the majority of cases having a broad 
band of ornament round the top. The drinking-cups 
are more elaborately decorated than any other class of 
sepulchral pottery, although the food-vessels are also 
nearly as ornate. 

The artistic quality of the decoration varies in different 
parts of Great Britain. Some of the most beautiful 
examples come from localities where there was a great 
mixture of aboriginal blood with that of the Celtic 
invaders, and it is not unlikely that the infusion of new 
blood may have had something to do with the excellence 
of the art. 

The chevron, although it was more highly developed 
as a decorative art-motive in the Bronze Age than at 
any other period, was not unknown to the Neolithic 
inhabitants of Great Britain, and it is more than 
probable that the Goidelic Celts got the idea from 
them. Several shallow vessels with a band of chevron 
ornament round the rim were found in the chambered 
cairn at Unstan,^ Orkney, which is of the later Stone 
Age. This particular chevron pattern occurs frequently 
in the Bronze Age. Each of the triangles formed by 
the chevron is filled in with hatched lines running 
diagonally, but alternately in directions at right angles 
to each other (Fig. 4, r/, p. 30). The pattern had no 
doubt a structural origin, and was suggested by lashing 
of the description used for the hafting of stone axes, or 
by some similar bandaging of cords. '^ 

^ Dr. J. Anderson's Scotland in Pagan Times: Bronze and Stone Ages, 
p. 294. 

- Prof. A. C. Haddon's Evolxitio)i in Art, p. 87. 



A similar chevron pattern is to be seen on a bowl 
from the Dolmen du Port-Blanc, Saint Pierre, Quiberon, 
Morbihan, Brittany.^ Possibly this may be the survival 
of a strengthening band of basketwork round the vessel. 

Bronze Spear-heads ornamented with rows of dots 

In the Museum of the Royal Irish Academy, Dublin 

The decoration of the bronze implements, gold lunulae, 
and jet necklaces of the Bronze Age corresponds very 
nearly with that of the sepulchral pottery. All the 

^ Paul du Chatellier's La Poterie aux Epoques pr^historique et Gauloise, 

pi. 12, fig". 12. 



designs are founded upon the chevron, and the only 
differences are in the methods of execution. On the 
objects of metal the patterns are produced by the 

Gold Lunula from Killarney 
Now in the Museum of the Royal Irish Academy, Dublin 

hammer, punch, and graver,^ and on the flat jet beads 
of the necklaces by a borer. 

The bronze implements most frequently decorated 
are celts and razors, and more rarely dagger-blades and 

1 Sir W. Wylde's Catal Mus. R. I. A., p. 388. 


Of the three classes of bronze celts, namely,^ 
(i) Flat celts, 

(2) Winged and flanged celts, 

(3) Socketed celts, 

it is only the first two that are decorated with chevron 
patterns in the same way as the sepulchral pottery. 
The socketed celts, which are later than the others, are 
ornamented with concentric circles resembling those on 
certain Gaulish terra-cotta figures.-^ 

On some of the bronze spear-heads in the Museum 
of the Royal Irish Academy the ornament consists of 
lines of small dots. The dotted patterns in the Irish 
MSS. of the Christian period may possibly be traced 
to this source. 

The greatest number of gold lunulce, most of which 
exhibit the characteristic chevron-motive decoration of 
the Bronze Age, have been found in Ireland. Dr. W. 
Frazer has compiled a list of known examples, which 
will be found in the Journal of the Royal Society of 
Antiquaries of IrelandJ^ The numbers are as follows : — 
Museum of the Royal Irish Academy . 32 

British Museum 
Edinburgh Museum 
Belfast Museum 
Private Collections . 
Present owners unknown 
Found in France 




The decoration consists of very fine lines executed 
with chisel-edged punches,"* and it is concentrated on 
the edges and the two horns of the crescent, the broad 

^ Early Man in Britain^ p. 350 ; and British Mnseuui Bronze Age 
Guide, p. 40. 

'^ Archceologia Camhrensis, ser, 3, vol. xiv. , p. 308. 

'^ 5th sen, vol. vii. , p. 41. 

^ Sir W. Wilde's Catal. of Antiquities of Gold in Mus. R. I. A., p. 10. 


part of the crescent in the middle being quite plain, as 
will be seen in the specimen illustrated on page 40 
from Killarney, now in the Museum of the Royal 
Irish Academy in Dublin. 

The lunulce were probably worn as head-dresses or 
else round the neck, and the contrast between the large 
expanse of burnished gold and the delicately engraved 
patterns must have been very effective when seen flash- 
ing in the bright sunlight. 

Some of the finest examples of jet necklaces have 
been found with Bronze Age burials in Scotland, as at 
Balcalk, Forfarshire; Tayfield, Fife; Torrish, Suther- 
landshire;^Assynt, Ross-shire;'- Melfort, and Argyllshire. 
They have also been found occasionally in England, as 
at Middleton Moor,-^ Derbyshire. 

The beads of which the necklaces are composed are 
of three different shapes, ovoid, flat triangular plates, 
and four-sided flat plates. The flat beads are decorated 
with chevrons, triangles, and lozenges produced by 
rows of dots. Here again we have an instance of a 
kind of decoration which survived in Christian times. 

The last class of remains exhibiting Bronze Age 
decoration are the sculptured rocks and stones. Some 
of the carvings are found on natural rock surfaces and 
boulders ; others on such megalithic monuments as 
stone circles, dolmens, and chambered cairns ; whilst 
numerous examples are on the slabs forming the covers 
or sides of sepulchral cists. 

Although the megalithic structures called by the late 
Mr. James Ferguson ''rude stone monuments " un- 

^ Dr. J. Anderson's Scotlajid in Pagan Times : Bronze and Stone Ages., 
PP- 53' 55' and 56. 

- Daniel Wilson's Prehistoric An)ials of Scotlatid. 
•' Bateman's Ten Years' Diggings., p. 25, 


doubtedly belong as a class to the Neolithic period, 
yet some of them exhibit decorative forms which are 
characteristic of the Bronze Age. This suggests the 
interesting speculation whether the ornamental patterns 
used by the Celts in the Bronze Age may not have been 
to a large extent borrowed from the Neolithic aborigines, 
and also whether the absorption of the Iberian peoples 
by the conquering Goidels may not have had a stimu- 
lating effect on decorative art. 

However this may be, it is a curious fact that the best 
specimens of Bronze Age ornament sculptured on stone 
exist in the Co. Meath, in Ireland, where such an ad- 
mixture of race would be most likely to occur, and the 
type of monument on which the carvings are found 
belongs to the Neolithic period. In Ireland, therefore, 
either the erection of dolmens, chambered cairns, and 
other similar structures must have survived during the 
Bronze Age, or else the characteristic patterns of the 
Bronze Age must have been derived from a Neolithic 

The wonderful series of chambered cairns at New- 
grange, near Drogheda, and at Sliabh na Calliaghe, 
near Oldcastle, both in the Co. Meath, have been well 
known to archaeologists for many years, but it is only 
quite recently that their decorative sculpture has been 
studied scientifically by Mr. George Coffey, m.r.i.a., 
the Curator of the Museum of the Royal Irish Academy 
in Dublin. The following account has been compiled 
chiefly from Mr. Coffey's admirable monographs on the 
subject, published in the Trmisactions of the Royal Irish 

The great prehistoric cemetery, which has been 
identified with the Brugh na Boinne mentioned in the 

^ Vol. XXX., p. I. 


Leabhar-na-h-Uidhri and in the Book of Ballymote, is 
situated five miles west of Drogheda, extending thence 
about three miles along the northern bank of the Boyne 
towards Slane. Amongst the most important of the 
sepulchral remains are the three great tumuli of Dowth, 
Newgrange, and Knowth, taking them in order from 
east to west. Two of the tumuli certainly contain 
chambers, access to which is gained by a passage 
leading from the exterior, and the third, judging from 
analogy, probably is also chambered. The Boyne 
tumuli are recorded in the Annals of Ulster to have 
been plundered by the Danes in a.d. 862. The chamber 
of the Dowth tumulus has been open since 1847 ; that 
of Newgrange since 1699, when it was first entered in 
modern times by Edward Lhuyd, the keeper of the 
Ashmolean Museum at Oxford ; and that of Knowth 
still remains to be explored. 

The sculptures at Newgrange are of such exceptional 
interest that it is desirable to give a brief description of 
the structure upon which they are found. The tumulus 
stands less than a quarter of a mile north-east of New- 
grange House, the Dowth tumulus being \\ to the 
north-east, and the Knowth tumulus three-quarters of a 
mile to the north-west. The Newgrange tumulus is 
surrounded by a circle of stones originally consisting 
of thirty-five upright monoliths, twelve of which may 
still be traced. Four of the standing stones near the 
entrance are from 6 to 7 feet in height, but the remainder 
are of smaller size. Between the circle and the base of 
the mound is a ditch and a rampart of loose stones. 
The tumulus is also of loose stones, surrounded at the 
base by a continuous curb of great slabs of stone from 
8 to 10 feet long, laid on edge, above which is a retaining 
wall of dry rubble 5 or 6 feet high. The tumulus is 

H : 


' I Ml? 




approximately circular in plan, 280 feet in diameter, 
and 44 feet high. The area occupied by the mound 
alone is at least an acre. The entrance to the passage 
leading to the chamber is on the S.E. side of the 
mound, and the passage runs in a N.W. direction. 
The chamber is not in the centre of the mound, but to 
the S.E. side of the centre. The plan of the passage 
and chamber is irregularly cruciform, the dimensions 
being as follows : — 

Length of passage . 

Length from end of passage to back of 

N.W. recess 
Average width of passage 
Width of chamber from back of N.E 

recess to back of S.W. recess 
Height of passage varies from 4 ft. 9 in. to 
Height of chamber 
Depth of N.E. recess 
N.W. „ 
S.W. ,, 



. 62 


. 18 

. 21 



• 19 


. 8 


• 7 


• 3 


The side walls of the passage and chamber are con- 
structed of tall upriglit stones, having the interstices 
filled in with rubble work. The passage is roofed over 
with single lintel stones. The roof of the chamber is 
in the form of an irregular six-sided truncated pyramid 
composed of stones corbelled out until they meet 
sufficiently near together at the top to be covered by 
a single slab. The floor was originally paved with 
carefully selected, water-rounded pebbles. These with 
equal originality and care have been removed by the 
Irish Board of Works, and placed in the bottom of the 
pit dug in front of the carved stone at the entrance. 

There are on the floor four rudely made shallow stone 



basins, one in each of the three recesses, and the fourth 
in the centre of the chamber. The one in the middle 
of the chamber was taken 
from the position it for- 
merly occupied on the top 
of the basin in the N.E. 
recess (where it was seen 
by Edward Lhuyd in 1699), 
and placed where it now 
is by the over-officious zeal 
of the Irish Board of 
Works. The large stones 
used in the construction of 
the chamber are of the 
lower Silurian grit of the 

The following stones of 
the Newgrange Tumulus 
are sculptured : — 

On exterior of Mound at Base, 

No. I. — Above entrance of 
passage leading to 

No. 2. — Front of entrance. 

No. 3. — Nearly in a line 
with axis of passage 
prolonged to cut cir- 
cumference of mound 
on N.W. side. 

No. 4. — N. side of monud. 

In Passage, '■■):...,'■■ ■^. ]:,.:■■:■■■■ 

N.E. side — twenty-one uprights — Nos. 3, 12, 18, and 21 

sculptured, counting from entrance inwards. 
S.W. side — twenty uprights — Nos. 10, 11, 17, and 20 

sculptured, counting from entrance inwards. 



In Chmiiber, 

Seventeen uprights — Nos. 2, 3, 4, 10, and 16 sculptured, 
commencing- at end of passage S.W. side, and counting 
round from right to left. Nos. 2, 3, and 4 are in S.W. 
recess, where there is also a horizontal stone above 
No. 3 sculptured. No. 10 forms the N.E. jamb of the 
N.W. recess. No. 16 forms the S.E. jamb of the 
N.E. recess, which has also a sculptured roofing-stone. 
The horizontal lintel-stone over the opening of the 
passage into the chamber is sculptured. 

Analysing the sculptured decoration of the New- 
grange tumulus, we find it to consist partly of chevron 

patterns and chevron deri- 
vatives (such as combina- 
tions of the triangle and 
lozenge), and partly of 
spiral ornament, together 
with a few designs formed 
of circles grouped round 
a lozenge, and some cups 
and rings. The chevron 
patterns have already been 
noticed on sepulchral urns, 
bronze implements, and 
jet necklaces of Great 
Britain, and concentric 

Spiral Ornament at Newg-range, 
Co. Meath. 

Scale i linear 

circles on socketed bronze 
celts, but spiral ornament 
is conspicuous by its absence on any of these classes 
of objects. Spirals are only known to occur on 
sculptured stones and rock-surfaces in Great Britain, 
and on a few of the remarkable stone balls with knobs 
found in Scotland. The following examples have been 
recorded : — 



Maughanby (Stone circle surrounding" cist under tumulus). 

Old Parks, Kirkoswald (Upright slab under tumulus). 
Lancashire. Calderstones, near Liverpool (Stone Circle). 
Northumberland. Morwick (Rock surface). 

Lilburn Hill Farm (Slabs of stone found in grave). 


Merionethshire. Llanbedr (Slab of stone found near hut- 
circles, now in Llanbedr churchyard). 


Orkney. Eday (Stone in Pict's House, nov^^ in the Edin- 
burgh Museum). 
Firth (Slab of stone, now in the Edinburgh Museum). 
Elginshire. Strypes (Standing stone). 

Elgin (Stone ball). 
Aberdeenshire. Towie (Stone ball, now in the Edinburgh 
Lumphanan (Stone ball, now in collection of Hugh W. 
Young, Esq., f.s.a. , Scot.). 
Argyllshire. Achnabreac (Sculptured rock-surface). 
Ayrshire. Coilsfield (Cist-cover). 

Blackshaw (Rock-surface). 
Peeblesshire. La Mancha (Slab of stone, now in the Edin- 
burgh Museum). 
Wigtonshire. Camerot Muir, Kirkdale (Standing stone). 
Dumfriesshire. Hollows Tower, Eskdale (Door-sill). 


Co. Meath. Newgrange (Chambered Cairn). 
Dowth (Chambered Cairn). 
Loughcrew (Chambered Cairn). 
King's Mountain (Chambered Cairn). 


Co. Louth. Killing- Hill, Dundalk (Sepulchral Chamber). 
Co. Tyrone. Knockmany (Chambered Cairn). 
Co. Fermanagh. Castle Archdall (Sepulchral Chamber). 
Co. Donegal. Glencolumbkille (Sepulchral Chamber). 

Spiral ornament is as conspicuously absent on the 
implements and objects of the Bronze Age in Gaul as 
in Britain. It is, then, to Scandinavia that we must 
look for the origin of the Bronze Age spirals found in 
this country. 

In the museums at Copenhagen, Stockholm, and 
Christiania, may be seen splendid specimens of bronze 
axes, sword-hilts, and personal ornaments exhibiting 
spiral decoration in the greatest perfection. 
These are fully illustrated in A. P. Madsen's 
monograph on the Bronze Age, in the works 
of O. Montelius and J. H. A. Worsaae, 
and in the Transactions of the various 
archseoloefical societies in Sweden and ^ . ,^ 

^ Spiral Ornament 

Denmark. on Bronze Axe- 

The spirals with which the objects of the head from 
Bronze Age in Scandinavia are decorated enmar 

are generally arranged with their centres at equal 
distances apart, and connected together by S- or C- 
shaped curves, the former being the most common. 

When spirals are arranged in a single row, the 
problem of how to connect the whole together so 
as to form a continuous running pattern does not 
present much difficulty, but if it is required to cover 
a large surface with spirals in groups of three or of four, 
all properly connected, the solution is not so easy as it 
appears at first sight. Both the metal-workers who 
made the Scandinavian bronze implements, and the 
artist who designed the sculptured decoration of the 
Newgrange tumulus, seem to have been unable to 



master the method of arranging the S- and C-shaped 
connections of the spirals in proper order,^ so as to be 
capable of extension in every direction over a surface 
of any required size. The difficulty was got over by 
a most ingenious artifice, as Mr. George Coffey was 

Bronze Axe-head with Spiral Ornament from Sweden 

the first to point out in his monograph on "Newgrange, 
Dowth, and Knowth " in the Transactions of the Royal 
Irish Academy (vol. xxx., p. 25). When the spirals 

^ That is to say, the way of placing the centres of the spirals in 
relation to each other, and of determining- how many S- or C-shaped 
curves should run to each centre. 


are not arranged and connected together in accordance 
with the requirements of geometry, some of the bands 
which compose the ornament have loose ends, i.e. run to 
nowhere. The question was how to dispose of the 
loose ends so as to deceive the eye and give the appear- 
ance of a continuous pattern. It was effected very 
simply by carrying the loose ends right round one or 
more of the other spirals so as to enclose them. Good 
instances of this occur on the sculptured slabs at 
Newgrange (p. 48), and on the carved stone ball from 
Towie, Aberdeenshire, now in the Edinburgh Museum 
of Antiquities. 

Mr. G. Coffey's theory, in which we feel inclined to 
agree, is that the spiral motive came to Ireland from 
Scandinavia across Scotland and the north of England. 
Both the geographical distribution of spirals sculptured 
on stone in Great Britain, and the fact that the same 
imperfect method of connecting the spirals together for 
all-over surface treatment is found in Ireland, Scotland, 
and Scandinavia certainly lend support to this view. 

It is now generally admitted by archaeologists that 
the spiral decoration of the Bronze Age in Scandinavia 
is of Mycenaean origin; and the clearest possible proof 
is furnished by an associated spiral and lotus motive 
design upon a bronze celt from AarhojV near Aalborg, 
Jutland, which finds an exact parallel in the ornament 
upon a gold pectoral from Myceuce.-^ 

The Mycenaean spiral decoration has furthermore 
been clearly proved by Mr. Goodyear in his Graviviar 
of the Lotus to have been borrowed from ancient 
Egypt ; the best instance of the transference of a 
spiral and lotus motive pattern from Egypt to the 

^ Mdmoires de la Socidtd Royale des Antiquaires du Nord, 1887, p. 259. 
^ Perrot and Chipiez's Art in Primitive Greece^ vol. i., p. 323. 


JEgea,n being the sculptured ceiling of the beehive 
tomb at Orchomenos. In Egypt, the spiral is found 
by itself forming a continuous running border on the 
scarabs of Usertesen I.^ (Twelfth Dynasty, B.C. 2758- 
2714), and combined with the lotus on a scarab at Turin^ 
of the same period. The best examples of the use of 
the spiral as continuous surface ornament are to be seen 
on the ceilings of Egyptian tombs of the Eighteenth 
Dynasty (b.c. 1633-1500).^ 

The spiral motive thus was most flourishing in 
Egypt from the Twelfth Dynasty to the Eighteenth, say 
from B.C. 2758-1700.^ After that it found its way to 
the ^gean, perhaps as early as 1400 B.c.,^ and thence 
to Hungary, Scandinavia, and Great Britain. 

The chambered tumuli at Dowth, on the Boyne, and 
Loughcrew, near Oldcastle, Co. Meath, resemble the 
Newgrange tumulus in plan and construction, but 
the sculptures upon the stones of the chambers and 
passages are not so obviously of Bronze Age type as 
those at Newgrange. The designs seem to be more 
symbolical than ornamental, and from the frequent 
occurrence of star- and wheel-shaped designs may have 
to do with sun-worship. The Loughcrew tumuli and 
their sculptures have been very fully described by 
Mr. E. A. Con well, in the Proceedings of the Royal 
Irish Academy (vol. ix., p. 355 ; and 2nd sen, vol. ii., 
p. 72); by Mr. George Coffey, in the Transactions of the 
same society (vol. xxxi., p. 23) ; and by Dr. W. Frazer, 
in the Proceedings of the Society of Antiquaries of 
Scotland {yo\. xxvi., p. 294). 

^ Flinders Petrie, Egyptian Decorative Aj't, p. 21. 

'^ Ibid., p. 22. 

3 Prisse d' Avenues , Histoire de V Art Egyptien apres les Monuments. 

^ Flinders Petrie, Decorative Art in Egypt, p. 28. 

^ Journal of Hellenic Studies, vol. xii. , p. 203. 


A certain proportion of the sepulchral cists of the 
Bronze Age in Great Britain exhibit symbolical or 
decorative designs. The following is a list of the 
examples which have been recorded : — 

Ross-SHiRE, Bakerhill. 
Argyllshire. Kilmartin. 

Clackmannan. Tillycoultry. 
Linlithgowshire. Caerlowrie. 

Craigie Wood. 
Lanarkshire. Carnwath. 
Ayrshire. Coilsfield. 
Cumberland. Aspatria. 

Redlands, near Penrith. 
Northumberland. Ford West Field. 
Yorkshire. Bernaldby Moor. 
Co. Tyrone. Seskin. 

The sculpture is usually on the cover-stone of the 
cist, but in the case of the examples at Kilmartin and 
at Carnban it is on the vertical end slabs. 

The sculptured designs consist of cups and rings, 
concentric circles, lozenges, triangles, axe-heads, curved 
meandering lines, and a few patterns composed of 
straight lines. The carvings show the same pick-marks 
that were observed at Newgrange. 

The axe-heads on the end slab of the cist at Kilmartin ^ 
are of the wedge shape common in the early Bronze 
Age. Like the stone axes and axe-heads sculptured 
on the dolmens of Brittany, they probably have a 
symbolical meaning connected with the worship of 
some axe-bearing deity such as Zeus. 

^ Jour. Brit. Archceol. Assoc, vol. 36, p. 146. 


The designs, composed of triangles alternately 
covered with dots and left plain, which occur on the 
cist-cover from Carnwath,^ we have already seen 
sculptured at Newgrange and engraved on bronze axes 
and jet necklaces. The grouped circles on the cist- 
cover from Craigie Wood- may also be compared with 
those on the slabs in the Newgrange tumulus, on the 
stone ball from Towie in the Edinburgh Museum, and 
on the chalk drums from Folkton in the British Museum. 

In three cases (viz. at Coilsfield,^ Carnwath, and 
Tillycoultry)^ elaborately ornamented urns of the food- 
vessel type have been found in the sculptured cists, 
thus clearly proving the period to which the cists 

Sometimes slabs of stone sculptured with cup-marks, 
cups and rings, and spirals, have been found associated 
with Bronze Age burials, although not forming parts 
of a cist. One of the most remarkable discoveries of 
this kind was made at Old Parks,'' near Kirk Oswald, 
Cumberland. In 1894 ^ barrow composed of loose 
stones, 80 feet in diameter and 4 feet high, was 
opened by the late Chancellor R. S. Ferguson, f.s.a., 
and when the mound was removed a row of five slabs 
fixed upright in the ground was disclosed. The stones 
were in a line pointing north and south, cutting the site 
of the mound into two halves, and three of them are 
sculptured with spirals. As many as thirty-two deposits 
of burnt bones were found in holes scooped out of the 
natural surface of the ground, together with two orna- 

^ Pmc. Soc. Ant. Scot., vol. x., p. 62. 

■■^ Ibid., vol. vi., Appendix, p. 28. 

'•^ Ibid., vol. vi., Appefidix, p. 27. 

^ Ibid., vol. xxix. , p. 190. 

•^ Cuinb. and West. Ant. Soc. Trans., vol. xili., p. 389. 


mented urns of incense-cup form, fragments of several 
other urns, and a necklace of cannel-coal beads. 

A slab of stone sculptured with spirals and concentric 
circles was found in 1883 on Lilburn Hill^ Farm near 
Wooler, Northumberland, associated with seven deposits 
of burnt bones buried in small circular pits. 

Stones sculptured with cups, or cups and rings, have 
been found either as cover-stones of urns or associated 
with burials in round barrows at the following places: — 

Northumberland. Ingoe. 

Black Hedon. 

Kirk Whelpington. 
Cumberland. Maughanby. 
Yorkshire. Kilburn. 

Ayton Moor. 

Claughton Moor. 

Wykeham Moor. 
Derbyshire. Elkstone. 

Staffordshire. Stanton. 
Dorsetshire. Came Down. 
Sutherlandshire. Dornoch Links. 
Aberdeenshire. Greenloan, Cabrach. 

A link between the art of the Bronze Age in Britain 
and the art of Mycense is afforded by a rock-sculpture 
at Ilkley," Yorkshire, which takes the form of a curved 
swastika. It belongs to a peculiar class of patterns 
composed of winding bands and small bosses or dots, 
of which there are numerous examples in the Scandi- 
navian^ and Mycenaean^ metal-work. Perhaps some of 

^ ArchcEol. ^liajia, sen 2, vol. x. , p. 220. 

'^ Jour. Brit Archceol. Assoc, vol. xxxv., p. 18. 

^ A. P. Madsen's A^itiquitds prdhistoriqiies du Danemarh. 

^ Schlieman's Mycence, pp. 166, 167,169, 264, and 265. 


the Late-Celtic designs, in which the arrangement of the 
long sweeping S- and C-shaped curves is governed by 
the position of circular bosses they connect, may be 
descended from the winding-band patterns of the 
Mycen^an period. For instance, the designs on the 
enamelled handles of the bowl found at Barlaston, 

Winding' Band (curved Swastika), sculptured in rock 

near Ilkley, Yorkshire 

Scale I linear 

Staffordshire, and on the Ilkley rock-sculpture have 
obvious points in common, both being founded on the 
curved swastika. 

There are in different parts of Great Britain a great 
number of rocks and boulders sculptured with cups, 
generally surrounded by concentric rings, and often 
having a radial groove leading from the cup outwards. 













The best-known instances are at Ilkley in Yorkshire, 
Wooler in Northumberland, the district on the east 
side of Kirkcudbright Bay between Kirkcudbright and 
the Solway Firth, and Lochgilphead and Kilmartin in 

Cup-and-ring Sculptures on rock at Ilkley, Yorkshire 

Scale 5V linear 

Argyllshire. In a few cases the cup-and-ring sculp- 
tures are associated with the wheel-symbol, as at 
Mevagh, Co. Donegal, and at Cochno, Dumbarton- 
shire. Such sculptures are more likely to be symbolical 
than decorative, but it would take us too far afield to 


discuss their meaning here. Those who wish to pursue 
the subject further may with advantage consult Sir 
James Simpson's valuable paper on "Ancient Sculp- 
turings of Cups and Concentric Rings," forming the 
Appendix to vol. vi. of the Proc. Soc. Ant, Scot. 

The sculptured rock-surfaces of Great Britain in 
some respects resemble the " Hallristninger " on the 
west coast of Sweden. The cup and ring, the wheel- 
symbol, and the curved swastika are common to both, 
but the Swedish sculptures are much more elaborate 
and include figure-subjects, ships, animals, etc. The 
age of some of the sculptures is indicated by the 
characteristic shape of the axes (evidently of bronze) 
held by the figures, and by the fact that the same set of 
symbols which occur on the rocks are also to be seen 
on the engraved knives of the Bronze Age found in 
Scandinavia. The Swedish rock-sculptures are fully 
described and illustrated in L. Baltzer's H'dllnstnmgar 
frail Bohiislan^ A. Holmberg's Skandinaviens H'dllrist- 
ninger, and the Memoires of the International Congress 
of Prehistoric Archaeology at Stockholm. 

Summing up the results of our investigations, we 
find that the peculiarities in the Pagan Celtic art of 
the Bronze Age which were transmitted to the Pagan 
and Christian styles of the Early Iron Age are as 
follows : — 

(i) The use of the closely coiled spiral. 

(2) The use of rows of dots. 

(3) The use of diagonal lines in preference to those run- 

ning horizontally or vertically. 

(4) The use of designs founded on the curved swastika. 

Of all these the spiral decorative motive is by far the 
most important, as we shall see in a subsequent chapter. 




IN a previous chapter we pointed out the difference 
between the Q and the P Celts, the former being 
Goidels in the Bronze Age, and the latter Brythons 
in the Iron Age, when they first arrived in Britain. 
We will now proceed to consider the nature of the 
culture introduced with the use of iron into this country 
from Gaul by the Belgic or Brythonic Celts. 


A great variety of circumstances have led to the 
discovery of objects of the Early Iron Age. Where 
they have not been buried at any great depth beneath 
the surface of the ground, the plough^ has frequently 
been the means of bringing them to light. The making 
of roads^ and railways,^ drainage of land for agricul- 
tural purposes,* military fortifications,'' quarrying*^ and 

^ As at Polden Hill, Somersetshire. 
^ As at Birdlip, Gloucestershire. 

^ As at cutting's near Bedford and between Denbigh and Corwen. 
4 As at Westhall, Suffolk. 
^ As at Mount Batten, near Plymouth. 
^ As at Hamdon Hill, Somersetshire. 



mining,^ have also had their share in helping the 
archceologist. A considerable number of antiquities 
which have found their way into the beds of rivers 
have been recovered in the course of dredging opera- 
tions for the improvement of inland navigation- and 
building of bridge foundations.^ Tumuli,^ camps, ^ 
caves/* sites of towns ^ and villages, '^ crannogs,^ etc., 
have yielded a plentiful harvest to the scientific ex- 
plorer. In some cases the denudation of the wind^^ or 
the erosion of the sea^^ has removed the covering of sand 
by which the traces of the ancient inhabitants have 
been concealed for centuries. The rabbit,^^ although 
the enemy of the farmer, sometimes becomes the friend 
of the antiquary by throwing up priceless relics of the 
past out of his burrow. Lastly, pure accident^^^ is now 

^ As at Hunsbury, near Northampton. 

'-' As in deepening- the Shannon, Thames, and Witham. 

^ As at Kirkby Thore, on the Eden, Westmoreland. 

^ As at Arras, Yorkshire. 

•'' As at Mount Caburn, near Lewes. 

•^ As at Settle, Yorkshire; Dcepdale, Derbyshire; and Kent's Cavern 
near Torquay. 

'• As at Great Chesters and Silchester. 

^ As at Glastonbury, Somersetshire. 

^ As at Lisnacrog-hera, Co. Antrim ; Strokestown, Co. Roscommon ; 
and Lochlee, Ayrshire. 

^^ As on the Culbin Sands, Elginshire, where in 1827 a sportsman 
having- lost his g-unflint, found a splendid Late-Celtic bronze armlet, 
whilst seeking- for another flint on the site of a Neolithic settlement covered 
with blown sand, except where denuded by the wind. 

^^ As at Hoylake, in Cheshire, where the encroachment of the sea on 
the portion of the coast lying- between the estuaries of the Dee and the 
Mersey washes out antiquities of every period from the submarine forest 
and the sandhills above it. 

^'^ A beautiful Late -Celtic bronze armlet was found at Stanhope, 
Peeblesshire, by the tenant of the farm, whilst searching- for a rabbit, 
under a larg-e flat stone on the hillside. 

^^ As in the case of the hoard of g-old objects of bullion value, amount- 
ing to £110, found at Shaw Hill, Peeblesshire, by a herd-boy who saw 
something glitter in the ground, and scraped out the torques and other 
relics with his foot. 


and then the agent by which the position of a long- 
forgotten hiding-place for valuables is made known. 


The general character of the finds of objects of the 
Early Iron Age is almost as varied as the circum- 
stances which have led to their recovery from oblivion, 
and they may be classified according to their nature, 
as follows : — 

(i) Sepulchral remains. 

(2) Remains found on inhabited or fortified sites. 

(3) Hoards of objects purposely concealed. 

(4) Objects accidentally lost. 

Sepulcln^al Remains. — The sepulchral deposits of the 
Early Iron Age differ greatly, both as regards the 
methods of burial adopted in each case, and the kind 
of grave-goods placed with the deceased. This is to be 
accounted for by a difference of time rather than area ; 
and it is only natural to find the Bronze and Iron Ages 
merging into one another, whilst towards the close of 
the Late-Celtic Roman and even Saxon influence began 
to be felt. 

Possibly the earliest sepulchral remains of the Late- 
Celtic period that have been found in England are 
the burials under mounds at Arras, on the Yorkshire 
Wolds, which were explored by the Rev. E. W. Stilling- 
fleet, D.D.,^ in 1815-17, and the Rev. Canon W. Green- 
well^ in 1876. The bodies were not cremated, as was 
generally the case in the Bronze Age, and also subse- 
quently during the Romano-British period ; but were 

^ Memoirs of the Meeting of the British A rchceological Institute held 
at York in 1846, p. 26. 

2 Greenwell's British Barrows, p. 454. 


buried in excavations in the chalk, and the place of 
sepulture marked by a tumulus. The so-called Queen's 
Barrow at Arras, when opened by the Rev. W. Stilling- 
fleet, was found to contain the skeleton of a female, 
with the feet gathered up, and the head to the north. 
The grave-goods consisted of one hundred glass beads, 
two bracelets, rings of gold and amber, and a pair of 

In another barrow at Arras, the Rev. W. Stillingfleet 
discovered the remains of a warrior resting on the 
smooth pavement of a circular excavation in the chalk, 
8 to 9 yards in diameter, and i foot 6 inches deep, lying 
on his back, with his arms crossed over the breast. He 
had been interred with his chariot, a pair of horses com- 
pletely harnessed, and two wild boars. 

A third barrow explored by the Rev. W. Stilling- 
fleet also covered the skeleton of a warrior with the 
remains of his martial equipment, consisting of the 
bosses of his shield, one wheel of his chariot, two of 
his horses' bridle-bits. Two wild boars' tusks (one of 
which was perforated with a square hole, and enclosed 
in a case of thin brass) were associated with this burial; 
indicating, perhaps, some religious or superstitious 
belief connected with this animal.^ 

A portion of the antiquities mentioned are now in 
York Museum, and the Rev. W. Stillingfleet's manu- 

^ A Late-Celtic boar's head of bronze was found at Liecheston, in 
Banffshire, in 1816 (see Dr. J. Anderson's Scotland in Pagan Times: 
Iroyi Age, p. 117). Three little bronze figures of boars^ from Hounslow, 
now in the British Museum, are illustrated in the Proc. Soc. Ant. Loud. 
(2nd ser., vol. iii. , p. 90); and the splendid bronze shield from the Thames 
at Battersea, in the same collection, has a boar represented upon it (see 
Kemble's Horce Ferales, pi. 14). The boar also occurs on one of the 
Scotch symbol-bearing- slabs at Knock-na-Gael, near Inverness (see 
Stuart's Sculptured Stones of Scotland, vol. i., pi. 38). For a boar on a 
helmet, see account of Benty Grange tumulus on p. 67. 


script notes on his diggings in 1815-17 are preserved in 
the Library of the York Philosophical Institute. 

The barrow at Arras, opened by the Rev. Canon 
W. Greenwell, covered a circular grave, 12 feet in 
diameter, sunk in the chalk to a depth of 3 feet, on the 
floor of which was laid the skeleton of a woman, resting 
on the left side, with her left hand up to the face, and 
the head to the west. Two tame pigs were buried with 
the deceased, and the grave-goods comprised an iron 
mirror, a bronze harness-ring, a pair of iron chariot- 
wheels, two snaffle-bits, and what may have been a 

In 1875 Canon Greenwell explored a tumulus near 
Beverley, in Yorkshire, which yielded two chariot- 
wheels and a bridle-bit, but no human or other bones. 

In July, 1897, Mr. J. R. Mortimer, of Driffield, 
opened 16 out of a group of 178 barrows, called ''Danes' 
Graves," near Pockthorpe Hall, two miles west of 
Kilham, E. R. Yorkshire. ^ The burial-mounds were 
from 10 to 33 feet in diameter, and from i foot 3 inches 
to 3 feet 6 inches high, covering graves, either oval or 
oblong with rounded corners, about 7 feet long by 5 feet 
wide by 2 feet deep. All the bodies were unburnt and 
buried in the doubled-up attitude characteristic of the 
Neolithic period. A beautiful bronze pin, inlaid with 
shell, was associated with the skeleton of a female in 
a grave beneath the largest of the mounds, and in 
another were found two male skeletons buried with a 
chariot, the iron tyres of the wheels and the iron hoops 
of the naves of which still remained together with the 
two iron snaffle bridle-bits of the horses. The antiqui- 
ties derived from the '' Danes' Graves" are now in the 

^ Reliquary iov 1897, p. 224 ; Proc, Soc. Ant. Lond.., 2nd sen, vol. xvii., 
p. 119. 



Museum of the York Philosophical Society. The 
average breadth index of the skulls was '735. 

The burials just described bear a marked resemblance 
to those of Gaulish warriors at Berru^ and at Gorge- 
Meillet,- both in the Department of the Marne in France, 
and may have belonged to the Celtic tribe of the 
Parisi, who gave their name to Paris in Gaul, and who 
colonised or conquered parts of Yorkshire. 

Canon Greenwell describes the result of opening four 
barrows of the Early Iron Age in the parish of Cowlam,^ 
in Yorkshire, in all of which were found the skeletons 
of females, laid on the natural surface of the ground, 
resting on the left side, with the hands up to the face, 
and the head to the north-east. The grave-goods from 
the first barrow consisted of a bronze armlet, a bronze 
fibula with an iron pin, and seventy exquisite blue 
glass beads ; and from the second, of an ornamental 
armlet. From the remaining two barrows only frag- 
ments of pottery were obtained. 

Mr. J. R. Mortimer explored a grave dug in the chalk, 
but without any mound above it, in 1868, a quarter of 
a mile north-east of Grimthorpe^ House, near Pockling- 
ton, in Yorkshire. It measured 4 feet 6 inches long, 
by 2 feet 9 inches wide, by 4 feet deep, and contained 
the skeleton of a young man, placed on the floor of the 
grave, resting partly on the back, with the knees and 
head inclined to the left side, the lower extremities 
drawn up, the hands on the breast, and the head to the 

^ A. Bertrand, Archdologie Celtique et Gauloise, 2nd ed., 1889, p. 356. 

^ E. Fourdrignier, Double Sdp^dture Gauloise de la Gorge-Meillet. 

■^ British Barrows, p. 208, Nos. li. to liv. The results of the explora- 
tion are now in the British Museum. The bronze objects are engraved 
in Sir J. Evans' Ancient Bronze Implements^ pp. 387, 388, and 400. 

^ Reliquary, vol. ix. , p. 180, and LI. Jevvitt's Grave-Mou7ids and their 
Contents, pp. 237 and 263. 


south. Associated with the burial were sixteen bone 
implements, a sword-sheath, the umbo of a shield, a 
disc of bronze with repousse ornament, and bits of 
rude pottery. 

The number of burials of the Early Iron Age that 
have been found in Great Britain is extremely small as 
compared with those of the Ages of Stone and Bronze. 
This would seem to indicate that the period between 
the introduction of iron into this country and the com- 
mencement of the Roman occupation cannot have been 
very long ; and that if the new metal was brought in 
by a foreign invasion rather than by peaceful com- 
mercial intercourse, nothing like the extermination of 
the native inhabitants, who used bronze and cremated 
their dead, can have taken place. 

As we have seen, a large proportion of the sepulchral 
remains of the Early Iron Age have been derived from 
Yorkshire ; but other instances have come to light in 
Derbyshire, Staffordshire, Kent, Gloucestershire, Devon, 
and Cornwall. 

The Rev. Mr. Pegge has given an account in the 
Archceologia^ of the opening of a tumulus on Garratt's 
Piece, Middleton Common, Derbyshire, a mile and a 
half south-east of Arbelows, and ten miles south-east of 
Buxton. The body had been laid on the surface of the 
ground, lying east and west. With it were found one 
of the circular enamelled discs to which reference will 
be made subsequently; a shallow basin of thin brass, 
much broken and crushed ; and part of the iron umbo 
of a shield. 

At Benty Grange, in Derbyshire, eight miles south- 
east of Buxton, on the road to Ashbourne, and one mile 

1 Vol. ix., p. 189: letter read May 8th, 1788; and T. Bateman's 
Vestiges of the Antiquities of Derbyshire, p. 24. 


north-west of Arbelows, Mr. Thomas Bateman^ exca- 
vated a barrow, about 2 feet high, surrounded by a fosse. 
The body had all decayed, except the hair ; but in the 
spot where it had been deposited was a remarkable 
assemblage of relics, consisting of a leathern cup 
mounted with silver round the edge, and having wheel- 
or cross-shaped silver ornaments round the bowl; three 
circular enamelled discs of the same class as those from 
the Middleton Common tumulus previously described ; 
an iron helmet surmounted by the figure of a hog of 
iron with bronze eyes, having a small silver cross inlaid 
on the nasal; a buckle; fragments of chains, etc. This 
burial, presenting some Celtic characteristics, belongs to 
a late period, possibly even after the Roman occupation. 

Two Early Iron Age burials are recorded as having 
been discovered in Staffordshire, one at Alstonfield, the 
other at Barlaston. The barrow near Alstonfield, cajled 
Steep Lowe,'- was composed of loose stones, and was 
50 feet in diameter, and 15 feet high. The Iron Age 
interment was a secondary one, the tumulus having 
been made originally in the Bronze or Stone Age. 
The body was laid on its back ; and amongst the 
grave-goods were a spear-head, a lance-head, and a 
knife (all of iron), some fragments of a highly orna- 
mented drinking-cup, a stud of amber, and Roman 
coins of Constantine and Tetricus. 

The burial at Barlaston,^ unlike the one just described, 
was not in a mound, but in a grave, 7 feet long by 
2 feet wide, by 1 foot 3 inches deep, cut in the solid 
red sandstone rock. With the body were associated a 

^ Ten Years' Diggings^ p. 28. 

- Bateman's Vestiges of the Antiqtiities of Derbyshire^ p. 76. 
•' LI. Jewitt's Grave-Montids and their Contents, p. 258; and Archceologia, 
vol. Ivi., p. 44. 


J^. ir. Dugdali photo. ' y 



beautifully ornamented flat bronze ring of Late-Celtic 
character ; three circular, enamelled discs of the type 
found in the barrow on Middleton Moor; some fragments 
of a bronze bowl, which Mr. LI. Jewitt erroneously 
conjectured to have formed portions of a helmet ; and 
blades of an iron sword and knife. 

No discovery of sepulchral remains belonging to the 
Late-Celtic period surpasses in interest that made in 
1879, between Birdlip^ and Crickley, on the Cotteswold 
Hills, seven miles south-east of Gloucester, both on 
account of the completeness of the series of objects 
buried with the deceased, and the extreme beauty of 
some of them as works of art. 

Whilst repairing the road, Joseph Barnfield un- 
earthed three skeletons interred with the feet to the 
south, in graves protected by thin slabs of stone placed 
on edge. The central skeleton was that of a female, 
and those on each side males. The following grave- 
goods were associated with the female : a bronze bowl 
(laid on the face of the deceased) ; a silver fibula plated 
with gold ; a necklace consisting of thirteen amber 
beads, two jet beads, and one marble bead ; a tubular 
brass armlet ; a brass key-handle ; a bronze knife- 
handle ornamented with a beast's head, having small 
knobs at the ends of the horns ; and last, but not least, 
a superb bronze mirror. 

Another very similar find of skeletons in graves 
formed of stones placed on edge was made in 1833 at 
Trelan Bahow,^ in the parish of St. Keverne, in Corn- 
wall, ten miles south-east of Helston. With one of 

^ See John Bellows, in Trans, of Bristol and Gloucestershire Archceol. 
Sac, vol. v., p. 137. The objects found are now in the Gloucester 

'^ See J. Jope Rodg-ers in Archceol. Joiirn.^ vol. xxx., p. 267. 


the skeletons was a beautiful bronze mirror, now in the 
British Museum. 

Sepulchral deposits of the same period, which have 
also yielded mirrors, were brought to light in the course 
of military works at Mount Batten,^ near Plymouth, 
in the spring of 1865. The burials, however, in this 
case were not in stone-lined graves near the surface, 
but in pits from 4 feet to 4 feet 6 inches deep, excavated 
in the disintegrated rock. In addition to a bronze 
mirror and the handles of two others, the following 
objects were obtained : two jointed bronze armlets, two 
plain bronze armlets, four fibulas, three bronze rings, 
a bronze cup, an iron dagger, and a pair of shears, 
black pottery, and fragments of glass. Ancient British 
coins had been found previously at Mount Batten,^ 
indicating a settlement here, perhaps in the first 
century B.C. 

The exploration of the Late- Celtic urn-field at 
Aylesford,'^ in Kent, three miles north-west of Maid- 
stone, by Dr. Arthur Evans, has been the means of 
extending our knowledge of the art of this period in a 
most unexpected manner, and has supplied the missing 
links between the culture of Britain in the first three 
or four centuries B.C., and that of La Tene on the 
Continent, which in its turn can be shown to have been 
strongly influenced by the civilisation of the ancient 
Venetian country at the head of the Adriatic* The 
shape of the tall, cordoned, pedestalled vases, and 
other peculiarities of the pottery from Aylesford, were 

^ See J. Spence Bate in Archceologia , vol. xl., p. 500, 
'■^ Sir J. Evans' Ancient British Coins, pp. 72 and 106. 
^ Archceologia, vol. Hi., p. 315. 

^ Dr. Arthur J. Evans' third Rhind Lecture on the " Origins of Celtic 
Art," as reported in the Scotsman, December 14th, 1895. 


things entirely unknown to archaeologists previously, 
and enable a distinction now to be drawn between the 
fictile ware of the Late-Celtic period and that of the 
Romano-British period. The discovery also of bronze 
objects of Italo-Greek manufacture of the second cen- 
tury B.C., associated with Late-Celtic burials, clearly 
indicates that there must have been a much more inti- 
mate trade-intercourse between Britain and the southern 
parts of Europe, in pre-Roman times, than has hitherto 
been suspected. 

The Late-Celtic urn-field at Aylesford was uncovered 
in 1886, at Messrs. Silas Wagon and Son's gravel-pit, 
in the course of removing the surface earth which here 
overlies the old river-deposits to a depth of 3 feet or so. 
One of the first burial-pits which attracted attention 
was circular, and about 3 feet 6 inches deep, the sides 
and bottom being coated with a kind of chalky com- 
pound. In the pit were found a bronze sitiila^ or pail, 
splendidly ornamented with repousse work in the 
Late-Celtic style, and containing calcined bones ; an 
oenochoe, or wine-jug ; and patella, or shallow pan, of 
imported Italo-Greek fabric ; fragments of a second 
sititla ; a bronze fibula ; and fragments of pottery. 

From another grave, about i foot 6 inches deep, 
situated 200 yards north-west of Aylesford Church, 
was obtained a bronze-plated tankard with two handles, 
of the same class as the Trawsfynydd tankard, sur- 
rounded by a circle of five or six earthenware vases, 
one of these being the finest pedestalled urn collected 
from the site. All the antiquities from Aylesford are 
now in the British Museum. 

Remains of the Early Iron Age found on Inhabited 
or Fortified Sites. — Next in importance to the sepulchral 
remains, as affording indications of the culture of the 


Early Iron Age, come the remains derived from in- 
habited or from fortified sites. And it may be remarked 
in passing that it is impossible to separate the inhabited 
from the fortified sites, because in these early times the 
state of the country was so unsettled that no isolated place 
of residence, village or town, could afford to do without 
some means of defence, either natural or artificial. 

The inhabited site which bids fair to rival all others 
in the varied nature of the relics obtained from it, and 
the light they help to throw on the arts and industries 
of the Early Iron Age in Great Britain, is the Glaston- 
bury Marsh Village. As the explorations begun by 
Mr. Arthur Bulleid, f.s.a., in 1892 are still in progress, 
it would be premature to pass an opinion upon the finds 
until they are completely exhausted. For an account 
of what has been already discovered there, the reader 
is referred to Mr. Bulleid's paper on the subject, which 
appeared recently in the Proceedings of the Somerset- 
shire Arcliceological Society.^ A bronze bowl is there 
illustrated which seems to be of the same kind as those 
derived from the graves, but it is ornamented with 
raised bosses instead of with circular plaques of enamel. 
The handle of a mirror, like those from the graves, was 
also found at the Glastonbury Marsh Village in 1896. 

From the exploration of this settlement we have 
obtained a knowledge of the peaceful pursuits and 
methods of life of the Late-Celtic inhabitants, which 
could never have been derived from their sepulchral 
remains. We now know that they were expert potters, 
wood-carvers, coopers, and weavers,-^ applying the 

1 Vol. xl. (1893). 

- Ornamental weaving- was, no doubt, practised. Althoug-h we have 
no absolute proof of this, the La Tene helmet from Gorg-e-Meillet 
(Marne), previously mentioned, has a sort of swastika pattern upon it, 
sug-gestive of a textile origin. 


same beautiful flamboyant forms of decoration that are 
characteristic of the metal-work of the period to earthen- 
ware and wooden vessels. The long-handled weaving- 
combs, which are so well known in the Pictish towers, 
or brocks, of the north of Scotland, have been found 
here also. Amongst the iron implements was a bill- 
hook for lopping the branches of trees — a most useful 
appliance for clearing away undergrowth in forests, 
procuring firewood, and building wattled structures. 
Unbaked ovoid clay pellets have been dug up in 
hundreds. These were probably sling -stones, in- 
dicating that the inhabitants must have been expert 

The dwellings appear to have been circular or oval 
wattled huts, the rudeness of which stands out in 
marked contrast to the high artistic taste and technical 
skill of the inhabitants. 

A few of the crannogs of Scotland^ and Ireland,-^ 
whose structure is somewhat analogous to the Glaston- 
bury Marsh Village, have also yielded Late-Celtic 
objects, but not in such quantities as to give evidence 
of permanent occupation over a considerable period. 

Hunsbury,^ two miles south-west of Northampton, 
which has been called the English "La Tene," is a 
good example of a Late-Celtic oppidum. The camp is 
of oval shape in plan, measuring 560 feet by 445 feet, 
and defended by a single earthen rampart and ditch. 

^ At Lochlee and at Lochspouts, Ayrshire ; Dowalton, Wigtownshire ; 
and Hyndford, Lanarkshire (see Dr. R. Munro's Lake-Dwellings of 

^ Lisnacrog-hera and Craig-y warren, Co. Antrim ; Strokestown and 
Ardakillen, Co. Roscommon ; Lagore, Co. Meath ; and BalHnderry, Co. 
Westmeath (see Wood Martin's Lake-Dwellings of Ireland). 

^ See Sir Henry Dryden in Associated Architectural Societies' Reports, 
vol. xviii. (1885), p. 53. 


The area enclosed is about four acres. Between 1880 
and 1886 the whole of the interior was excavated to 
obtain ironstone, which lay in a bed 12 feet thick, at 
a depth of 7 feet 6 inches below the natural surface 
of the ground. 

In the course of the excavations about three hundred 
refuse-pits, averaging 5 feet in diameter, and dug in 
the soil overlying the ironstone, were discovered. 
Amongst the contents of the pits were two bronze 
sword-sheaths, one of them highly ornamented in the 
Late-Celtic style ;^ three fibulas, bridle-bits and cheek- 
pieces of bone, a chariot-wheel, iron saws, knives, 
spear-heads, etc. ; one hundred and fifty quernstones, 
reckoning the upper and lower stones separately ; 
eight spindle-whorls, long-handled weaving-combs, 
and pottery with Late-Celtic decoration. All these 
antiquities are now in the Northampton Museum. 

The camp on Mount Caburn, two miles south of 
Lewes, in Sussex, explored by General Pitt-Rivers'^ in 
1878, seems to have been an oppidum of the same class 
as that at Hunsbury, and the relics indicated the same 
kind of culture. The pits found at Mount Caburn were 
some of them oval, and others oblong, 5 to 7 feet in 
diameter, and 5 feet deep. The objects obtained from 
the pits included ornamental pottery, long-handled 
wearing-combs, an iron billhook like the one from the 
Glastonbury Marsh Village, and three ancient British 
tin coins. 

The fine collection of Late-Celtic horse-trappings, 
etc., now in the Duke of Northumberland's private 
Museum at Alnwick Castle, was found in 1844, in a 
pit about 5 feet deep, within an earthen entrenchment 

^ Engraved in the Archceologia, vol. Hi., p. 762. 
- Archceologia, vol. xlvi., p. 423. 


at Stanwick, in Yorkshire, seven miles north of Rich- 
mond. ^ 

A few Late-Celtic objects have been derived from 
Roman towns^and stations^ in England; and also from 
the weenis^'^ or underground houses, and the brochs^^ or 
Pictish towers of Scotland. 

The bone-caves which were the permanent habita- 
tions of Paleolithic and Neolithic man in Britain served 
as temporary places of refuge for the Brit-Welsh popu- 
lation during the troublous times immediately succeed- 
ing the Roman evacuation of this country. Gildas' 
account of the Britons leaving their houses and lands, 
and taking shelter in the mountains, forests and caves, 
whence they were able successfully to repel the inroads 
of the Picts and Scots, ^ is fully borne out by archaeolo- 
gical research.^ 

The principal caves which have yielded relics of this 
period are Kirkhead^ Cave in Lancashire; the Victoria,-^ 

^ Memoirs of the Meeting of the Archceological Institute of Great 
Britain and Ireland at York in 1846, p. 88; Dr. J. C. Bruce's Catalogue 
of the Antiquities at Alnwick, p. 38. 

'■^ As in Silchester. These have not been illustrated, but are to be seen 
in the Reading" Museum. 

^ As in ^sica (Great Chesters) {A rchceologia JEliana, 2nd sen, 
vol. xvii. , p. xxviii. ). 

^ As at Castle Newe, Aberdeenshire, and Grange of Conan, Forfar- 
shire (see Dr. J. Anderson's Scotland in Pagan Times: Iron Age, pp. 141 
and 160). 

° As at Okstrow and at Harray in Orkney {Ibid., pp. 219, 236}. 

^ Gildas, xvii.; Bede's EccL Hist., bk. i., chap. xiv. 

'^ Prof. W. Boyd Dawkins' Cave-Hunting, p. 106. 

^ Three miles south of Cartmel, on the shore of Morecambe Bay 
{Cave-Hunting, p. 125). 

^ A mile and a half north-east of Settle {Cave-Hutiting, p. 81 ; and 
H. Eckroyd Smith in Trans, of Hist. Soc. of Lancashire and Cheshire, 
vol. for i866, p. 199; and Roach Smith's Collectanea Antigua, vol. i., 
p. 67). 


Kelko,^and Dowkerbottom" Caves in Yorkshire; Poole's" 
Hole and the Deepdale^ Cave in Derbyshire ; Thor's^ 
Cave in Staffordshire ; and Kent's^ Cavern in Devon- 

The character of the antiquities derived from the 
caves does not differ materially from that of the remains 
from the crannogs and the oppida^ although a few 
things of peculiar form have been found in some of the 
caves, such as the spoon-shaped bone-pins from the 
Victoria and Dowkerbottom Caves, and the bone 
whistles from Thor's Cave. The fibulae from the 
Victoria and the Deepdale Caves are of remarkable 
beauty. Evidence of spinning is afforded by the long- 
handled comb from Thor's Cave, and the numerous 
spindle-whorls from others. The discovery of Roman 
coins and Samian ware indicate the period at which the 
Brit-Welsh sought refuge in these recesses of the rock. 

Hoards of Late-Celtic Objects purposely co7tcealed. — 
The horse-trappings found in an excavation at the 
bottom of one of several oblong pits, 7 feet long by 
3 feet wide by 4 feet deep, at Hagbourne HilF in 
Berkshire, two miles south of Didcot, seem to have 
been purposely hidden ; as also the horse-trappings 
which were discovered in the chink of the rock by 
quarrymen at Hamdon Hill^ in Somersetshire, five 

^ Overlooking Gigg-leswick, one mile north-west of Settle. 

- Between Kilnsey and Arncliffe, ten miles north-east of Settle {Proc. 
Geol. and Poly tech. Soc. of IV. Riding of Yorksh. for 1859, p. 45). 

•^ A mile south-west of Buxton {Cave-Huntifig, p. 126). 

■* Three miles south-east of Buxton {Derbyshire Archceol. Soc. Trans. ^ 
vol. xiii. , p. 196). 

^ Near Grindon, eight miles north-west of Ashbourne {Reliquary^ 
vol. vi., p. 201, and Tra?is. Midland Sci. Assoc, 1864-5, p. 1). 

^ One mile north-west of Torquay. There is a fragment of pottery, 
with Late-Celtic ornament upon it, from Kent's Cavern, in the British 

" ArchcBologia, vol. xvi. , p. 348. ^ Ibid.., vol. xxi., p. 39. 


miles west of Yeovil. Another instance of intentional 
concealment is afforded by the beautiful bronze mirror 
that was found, with other ornamental pieces of bronze, 
wrapped in a cloth, and covered by the upper stone of 
a quern, at Balmaclellan,^ two miles north-east of New 
Galloway, Kirkcudbrightshire. 

Late-Celtic Objects accidentally lost. — Besides the 
Late-Celtic objects which have been dropped by their 
original owners on dry land, and got covered with the 
soil and thus been preserved, it is remarkable in how 
many cases they have been lost whilst crossing or navi- 
gating rivers, especially the Thames,^ Witham,'^ Tyne,^ 
and Tweed. '^ 


The earliest native coinage of Britain belongs to the 
Iron Age, and dates from 200 to 150 B.C. Sir John 
Evans has collected together all the known facts relating 
to the numismatics of this period in his Coins of the 
Ancient Britons^ and gives excellent maps showing the 
geographical distribution of the finds. Prof. John 
Rhys, in his Celtic Britain (p. 19), says : — 

"The coinage of Britain had been modelled in the first 
instance after that of Gaul, which, in its turn, can be traced to 
the Phocaean Greeks of Massilia or Marseilles, through 
whom the continental Gauls became acquainted in the latter 
part of the fourth century before Christ with the gold stater 
of Philip II. of Macedon. This was a fine coin, weighing 

^ Dr. J. Anderson's Scotland in Pagan Times : Iron Age, p. 126. 

^ Shield {ArchcBologia, vol. xxiii. , p. 96) ; helmet (in the British 
Museum); fibula {^Proc. Soc. Ant. Scot., 2nd sen, vol. xv. , p. 191). 

^ ^\i\Q\di i^Archceologia , vol. xxiii., p. 96); sword-sheath (J. C. Bruce's 
Catal. of Alnwick Mus.) ; dag-g-ers (Kemble's Horce Ferales, pi. 17). 

* Fibulae {Illustrated Archceologist, vol. ii. , p. 157). 

^ Sword-sheath {Archceologia, vol. xlv., p. 45). 


133 grains, and having- on one side the head of Apollo 
wreathed with laurel, while the other showed a charioteer in 
a biga, with Philip's name underneath. It was imitated by 
the Gauls fairly well at first, but as it got further removed 
from the original in time and place, the figures degenerated 
into very curious and fantastic forms." 

Before the landing of Julius Csesar in Britain in 
55 B.C. the Cantii, the Dutoriges, the Catuvelauni, and 
the Trinovantes each had coinages of their own, but 
entirely devoid of lettering. The lettered coins begin 
with those of Commios, dating from a period some 
time before 30 B.C., after which come those of his three 
sons — Tincommios, Verica, and Eppillos. Prof. J. Rhys^ 
says : — 

"The coins of Commios, and some of the earlier ones of 
Tincommios, continued the degenerate imitations of the 
Macedonian stater without showing any Roman influence ; 
but it was not long after Augustus became emperor, in the 
year 30, that Tincommios copied the Latin formula, in which 
the former styled himself Augustus Divi Filius, or the son of 
his adoptive father, Julius Caesar, who had now got to be 
officially called Divus^ or the god. So Tincommios had 
inscribed on his money the legend Tiuc. Conuni F., or even 
shorter abbreviations, meaning Tincommios son of Commios ; 
and the grotesque traits derived from the stater soon 
disappear in favour of classical designs of various kinds, 
proving very distinctlv that the influence of Roman art was 
beginning to make its f felt in the south of Britain." 

The coins which have been assigned to the Dobunni 
(although their exact date, place of issue, and sequence 
are somewhat doubtful) belong to the series of the 
Macedonian stater, and show hardly any trace of 
Roman influence. Their probable date'- seems to be 
between a.d. i and 41. 

^ Celtic Britain, p. 25. - [bid., p. 35. 


There was no native British coinage either in Scotland, 
Wales, or Ireland, and in England the finds do not 
extend further north than Newcastle-upon-Tyne. The 
greater part of the finds lie to the south of a line drawn 
from Wroxeter to the Wash, and east of a line drawn 
from the same place to Exeter. The geographical 
distribution of the finds is clearly shown on the map 
given in Sir John Evans' Coins of the Ancient Britons, 


In the present state of our knowledge no very satis- 
factory deductions can be made from a study of the 
geographical distribution of the finds of this period, 
partly because the discoveries have been so imperfectly 
recorded (more especially in Ireland), and partly because 
a large number of sites which are probably Late-Celtic 
still remain unexplored. Another difficulty to be 
reckoned with is that it is only within the last few 
years that arch^ologists have been able to distinguish 
between what is purely Celtic and what is Romano- 
British. Indeed, in many cases, in the absence of coins 
or other evidence, it is quite impossible to determine 
whether particular finds are of pre-Roman, Roman, or 
even post-Roman, as the Late-Celtic style of decoration 
was in vogue throughout the whole of the Pagan Iron 
Age in Britain, and survived in remote districts after 
the introduction of Christianity. Then, again, the fact 
must not be lost sight of that the greater frequency of 
finds in some districts than others can be accounted 
for by their having been covered with lakes where cran- 
nogs could be easily constructed, or on the limestone 

1 A complete list of the finds as far as recorded is given in the 
Archceologia Cambrensis, 5th sen, vol. xiii., p. 321. 


formation, where rock-shelters and caves suitable for 
temporary places of refuge already existed. 

We will take the geographical distribution of the 
sepulchral remains first. It is most remarkable that 
up to the present no Late-Celtic burials have been 
recorded in Wales, Scotland, or Ireland, although finds 
of objects of the period have been frequent in all three 
of these countries. The earlier Bronze Age burial 
customs may, of course, have survived after the intro- 
duction of iron, or interments of the Iron Age may 
have passed unrecognised owing to the rapid decay of 
the metal implements accompanying the body. 

In England the greatest number of Late-Celtic burials 
have been found in the south-east corner of Yorkshire, 
near Beverley and Driffield. In most cases the tumuli 
covering the graves are in large groups, those at 
"Danes' Graves," near Kilham, numbering 178; those 
at Arras, near Market Weighton, 200 ; and those in 
Scorborough Park, near Beverley, 170. The people to 
whom these extensive cemeteries belonged probably 
invaded Britain from the Continent some few centuries 
before the Roman occupation, and landing at the 
mouth of the Humber, settled permanently on the east 
coast of Yorkshire. The people in question had long 
skulls, and buried their dead in a doubled-up attitude 
without cremation, which has suggested another less 
probable theory that they were the direct descendants 
of the more ancient Neolithic inhabitants of Yorkshire. 

In Derbyshire one undoubted Late-Celtic burial has 
been found^ and there are a few others which seem to 
belong to the transition towards the end of the Roman 
occupation or the beginning of the Saxon Pagan 
period. - 

^ In Deepdale. ^ At Benty Grange, and QH Middleton Moon 


In Kent^ and Devonshire''^ cemeteries containing a 
large number of graves have been brought to light. 

Isolated burials have been found in single localities 
in each of the counties of Stafford,^ Gloucester,^ Dorset,^ 
and Cornwall.^ 

The following lists show the geographical distribution 
of the inhabited or fortified sites of the Late-Celtic 
period in Great Britain : — 


Yorkshire. Dowkerbottom Hole, Arncliffe. 

Victoria Cave, Settle. 

Kelko Cave, Settle. 
Lancashire. Kirkhead Cave. 
Derbyshire. Thirst House, Deepdale. 

Poole's Hole, near Buxton. 
Staffordshire. Thor's Cave, Dovedale. 
Devonshire. Kent's Cavern, Torquay. 


Somersetshire. Glastonbury Marsh Village. 
Lanarkshire. Hyndford Crannog. 
Ayrshire. Lochlee Crannog. 

Lochspouts Crannog. 
Wigtownshire. Dowalton Crannog. 
Co. Antrim. Lisnacroghera Crannog. 

Craigywarren Crannog. 
Co. Roscommon. Strokestown Crannog. 

Ardakillen Crannog. 
Co. Westmeath. Ballinderry Crannog. 
Co. Meath. Lagore Crannog. 

1 At Aylesford. 2 ;^t Mount Batten. 

" At Barlaston. "^ At Birdlip. 

^ In the Isle of Portland. « At Trelan Bahow. 



Orkney. Broch of Harray. 

Broch of Okstrow. 
Caithness. Broch of Kettleburn. 
Selkirkshire. Broch of Torwoodlee. 


Aberdeenshire, Castle Newe. 
Forfarshire. Grange of Conan. 


Yorkshire. Stanwick. 

Northamptonshire. Hunsbury. 

Kent. Bigbury Camp. 

Sussex. Mount Caburn. 

Berkshire. Northfield Farm, Long- Wittenham. 

Dorsetshire. Hod Hill. 

Hambledown Hill. 

Maiden Castle. 

Somersetshire. Ham Hill. 
Nairnshire. Burghead. 
Perthshire. Abernethy. 
Ayrshire. Seamill Fort. 
Cardiganshire. Castell Nadolig-. 
Carnarvonshire. Treceiri. 


Northumberland. Great Chesters (^sica). 

Rising-ham (Habitancum). 
Westmoreland. Brough. 

Kirkby Thore. 
Lancashire. Ribchester. 
Yorkshire. New Malton. 
Northampton. Wellingborough. 
Surrey. Farley Heath. 
Hampshire. Silchester. 
Perthshire. Ardoch. 
Dumfries. Birrenswark. 


A study of the above lists discloses some interesting 
facts. It will be noticed that the caves are confined 
exclusively to the limestone districts of the counties of 
York, Lancaster, Derby, Stafford, and Devon. The 
lake-dwellings are found chiefly in the south-west of 
Scotland and the north-east and central part of Ireland, 
there being only one example in England and none in 
Wales. The brocks and weems (or underground houses) 
are Pictish structures, and therefore do not occur any- 
where except in Scotland, chiefly in its north-eastern 
counties. The Celtic oppida are most common in the 
south of England where the Belgic settlements pre- 
dominated, but there are a few examples in Scotland. 
Probably a more systematic examination of the hill- 
forts throughout Great Britain would show that those 
in which large areas are enclosed by double and triple 
ramparts of stone or earth ^ belong to the Late-Celtic 
period. At the present time practically nothing is 
known as to the age of the stone forts and earthen 
raths in Ireland or Wales. 

Most of the Romano-British fortified sites which 
have yielded works of art of Late-Celtic type, although 
executed under Roman influence, are in the south of 
Scotland or in the north of England, on or near the 
Wall of Hadrian, or along the lines of the military 
roads leading to it.'^ At Farley Heath, near Guild- 
ford, Surrey, numerous specimens of Kelto-Roman 
enamelled bronze objects have been found, and this 

^ I refer here to defensive works in which the whole of the summit of 
the hill is enclosed. These forts are usually of approximately oval shape, 
and follow the conformation of the hill. 

^ As, for instance, at Rising-ham (Habitancum) on the road going- 
north from the Wall into Scotland, and at Broug-h and Kirkby Thore 
on the road from York to Carlisle, which passes through upper Teesdale, 
and thence into the valley of the Eden. 


site would no doubt produce a plentiful harvest of 
antiquities of a similar nature if properly explored. 
The great difficulty, however, as we have already 
pointed out in dealing with the Romano-British sites, 
is to determine to what extent the style of the art of 
the objects found there can be shown to be definitely 
Celtic. In our lists we have only included sites from 
which have been procured antiquities exhibiting Celtic 
enamel and flamboyant ornament, or fibulae of known 
Celtic type. 

If to the sepulchral deposits and inhabited sites just 
described be added all the miscellaneous finds of objects 
accidentally lost or purposely concealed, it will be 
observed that there is hardly a single county through- 
out England, Wales, Scotland, and Ireland which 
cannot show one or two such Late-Celtic finds at least. 
Some counties are nevertheless richer than others,^ as, 
for instance, Aberdeen, Forfar, Perth, Ayr, Kirkcud- 
bright, and Dumfries, in Scotland ; Antrim, Meath, 
and Roscommon, in Ireland; Denbighshire, in Wales; 
and Northumberland, Durham, Yorkshire, Derbyshire, 
Suffolk, Middlesex, Kent, Gloucestershire, Oxford- 
shire, Berks, Wilts, Dorset, Hants, and Somerset, 
in England. 


The finds of objects of the Early Iron Age to which 
an approximate date can be assigned are as follows : — 

(i) Finds associated with burials of a particular kind. 

(2) Finds associated with objects of the Bronze Age. 

(3) Finds associated with objects of early Hallstatt type. 

^ i.e. have from four to ten localities where Late-Celtic finds have 
been made. 


(4) Finds associated with fibulae or other objects of 
La Tene type. 

(5) Finds of objects associated with imported articles of 
GrEeco-Italic fabric. 

(6) Finds of objects associated with Ancient British coins. 

(7) Finds of objects, (a) on Romano- British inhabited or 
fortified sites, [b) associated with Roman coins, and {c) 
associated with articles of Roman manufacture. 

There are at least three different methods of burial 
characteristic of the Late-Celtic period in Great Britain. 

(i) Uncremated burials in excavated graves beneath 
barrows in which the deceased is g"enerally found with his 
chariot and horses, as at Arras, Yorkshire. 

(2) Cremated burials in pits without any exterior mound, 
the ashes being* contained in cinerary urns and the burials 
in groups, as at Aylesford, Kent. 

(3) Uncremated burials in graves formed of slabs of stone 
placed on edge, without any exterior mound, as at Birdlip, 

The first class of burials correspond with those at 
Berru and Gorge-Meillet, Department of the Marne, 
and probably belong to the same period as these earlier 
Gaulish interments which, from the associated Greek 
and Etruscan relics,^ are known on the Continent to 
belong to the third, fourth, and fifth centuries B.C.- 

The second, or Aylesford urn field type of burial, is 
dated by associated vessels of Italo-Greek fabric at from 
200 to 150 B.C.^ 

Implements of the Bronze Age have been occasionally 
discovered with objects of Late-Celtic character, as at 

^ Such as the Grsecwyl bronze vase now in the Berne Museum ; the 
bronze cenochoe and Etruscan cup from Somme-Bionne (Marne) ; and 
the two-handled cup from Rodenbach, Bavaria (described and illustrated 
in A. Bertrand's Arch^ologie Celtique et Gauloise, pp. 328 to 347). 

" Arthur Evans in A rchceologia, vol. lii. , p. 72. 

^ Ibid., vol. lii., p. 66. 


Hagbourn Hill,^ Berks, where a Late-Celtic bridle-bit 
and harness- rings were associated with some small 
spear-heads and a socketed celt; and at Hounslow,^ 
Middlesex, where three figures of boars and two of 
other animals were found with celts and gouges of the 
Bronze Age. 

Up to the present time, no specimen has yet been 
found in this country of the great iron sword of Hall- 
statt type, with its massive ivory handle encrusted with 
amber.^ Of the smaller Hallstatt sword with an iron 
blade and a bronze handle, having antennae-like pro- 
jections at the top,^ one specimen from the Thames is 
to be seen in the British Museum, and there are about 
half a dozen others in the Museum of the Royal Irish 
Academy in Dublin. 

Tall vessels made of thin sheets of bronze riveted 
together and furnished with two round ring-handles at 
the top have been found in Ireland (at Montiaghs,^ 
Co. Armagh; and Dowris,^ King's Co.) and in Scotland 
(at Cardross,'^ Dumbartonshire) ; the form of these 
vessels shows that they are akin to the situlae of the late 
Hallstatt or early La Tene period on the Continent. 

In the instances where other objects have been asso- 
ciated either with the swords a antennes or situlae in 
Great Britain they have been of purely Bronze Age 
type, showing that the Hallstatt period on the Continent 
was earlier than the Late-Celtic period in this country. 

The forms of the fibulae associated with Late-Celtic 

^ Archceologia, vol. xvi. , p. 348. 

'-^ Proc. Soc. Ant. Loud., 2nd sen, vol. iii., p. 90. 

•^ A. Bertrand and S. Reinach's Les Celtes dans les Valldes du P6 et du 
Danube, p. 125. ^ Ibid., p. 85. 

•^ Jour. Royal Soc. Ant. Ireland., sen 5, vol. vii., p. 437. Another example 
found in Ireland is fig-ured in Sir W. Wylde's Catal. Mus. R.I. A., p. 531. 

^ Now in the British Museum. Evans' Ancient Bronze Implements., 
p. 410. ^ R. Munro's Prehistoric Scotland, p. 40. 


finds afford specially valuable evidence as to date. The 
pre-Roman, or La Tene, type of fibula was made in one 
piece on the same principle as the modern safety-pin, 
and therefore differed from the Roman Provincial harp, 
or bow-shaped fibula, in which the pin was separate 
from the back and worked on a hinge. Fibulae of the 
earlier kind have been found with Late-Celtic burials 
at Cowlam, Yorkshire ; Aylesford, Kent ; and Birdlip, 
Gloucestershire ; and on inhabited and fortified sites at 
Hod Hill, and Rotherley, Dorset. The fibulae from the 
Stamford Hill Cemetery, near Plymouth, and the Polden 
Hill hoard of horse-trappings belong to the later class. 
As the forms of the different fibulae will be discussed 
subsequently, no more need be said on the subject here. 

Ancient British coins have been found near the Late- 
Celtic cemeteries at Aylesford,^ Kent, and Stamford 
Hill,^ near Plymouth ; also within the fortifications of 
the Late-Celtic oppida at Mount Caburn,^ near Lewes, 
Sussex, and Hod Hill,^ near Blandford, Dorset. General 
Pitt-Rivers came across ancient British coins during 
his excavations on the site of the Romano-British 
village at Rotherley^ in Cranbourne Chase, Dorset, and 
numerous specimens (especially of the coins of Verica, 
one of the three sons of Commios) have turned up from 
time to time at Farley Heath, ^ near Guildford, Surrey. 

The Romano-British inhabited or fortified sites from 
which objects of Late-Celtic character have been derived, 
have already been specified. The following lists show 
the instances where Late-Celtic finds have been asso- 
ciated with Roman coins or with objects of Roman 
manufacture : — 

^ ArchcBologia^ vol. lii., p. 315. '-^ Ibid.., vol. xl. , p 500. 

- ^ Ibid., vol. xlvi. , p. 423. ^ Archceol. Jour.., vol. Ivii. , p. 52. 

^ Excavations in Cranbourne Chase, vol. ii., p. 188. 
^ F. Martin Tupper's Farley Heath, p. 10. 



Late-Celtic Finds associated with Roman Coins. 


Nature of Find. 

Date of Coins. 

Victoria Cave, Settle, 

Kelko Cave, Giggles- 
wick, Yorkshire 

Dowkerbottom Cave, 
Arncliffe, Yorkshire 

Kirkhead Cave, Cart- 
mel, Lancashire 

Poole's Cavern, 

Thirst House, Deep- 
dale, Derbyshire 

Thor's Cave, Stafford- 

Broch of Torwoodlee, 

Rotherley, Dorset 

Kirkby Thore, West- 
Hod Hill, Dorset 

yEsica, Northumber- 

Farley Heath 
Alstonfield, Stafford- 

Ham Hill, Somerset 

Westhall, Suffolk 
Chorley, Lancashire 

Inhabited Site 
Inhabited Site 
Inhabited Site 

Inhabited Site 
Inhabited Site 
Inhabited Site 

Inhabited Site 
Fortified Tower 
British Village 

Fortified Site 

Romano - British 

British Village 

Fortified Site 


Trajan to Con- 

stans, A.D. 98- 

(?) [353- 

Claudius Gothicus 
to Tetricus, a.d. 

Domitian, a.d. 81- 



Antoninus Pius to 
Gallienus, a.d. 


Vespasian, a.d. 

Trajan to Gallie- 
nus, A.D. 98-268 

Vespasian to Sev- 
erus, A.D. 69-211 

Augustus to Tra- 
jan, B.C. 27 — 

A.D. 117 

Mark Antony to 
Magnentius, B.C. 
32-A.D. 353 


Tetricus and Con- 

stantine, a.d. 

Valerian to Theo- 

dosius I., a.d. 

^ 3797395 
Hadrian, a.d. 117- 





Nature of Find. 

Date of Coins. 

Backworth, Northum- 


Antoninus Pius, 


A.D. 138-161 



Claudius, a.d. 41- 



Castlethorpe, Bucks 

Silver Armlet 

Verus,A.D. 161-169 

Late-Celtic Finds associated with 
Broch of Okstrow, Orkney 
Lochspouts . 
Settle, Yorkshire . 
Deepdale, Derbyshire . 
Thor's Cave, Staffordshire 
Isle of Portland 
Westhall, Suffolk . 

Sa7nian Ware. 

Nature of Find. 

Pictish Tower. 








Late-Celtic Fiiids 

associated with Objects of Roinan 


Nature of Late- 
Celtic Find. 

Class of Roman 

Dowalton, Wigtown- 

Stanhope, Peebles- 

Polden Hill, Somer- 

Stamford Hill, near 

Castlethorpe, Bucks 

^sica, Northumber- 

Hay Hill, Cambridge- 

Mount Bures, Essex 

Stanfordbury, Bed- 







Silver necklace, 


Six amphorae, 

glass, etc. 
Bronze jug, Sa- 

mian ware, etc. 




HE materials available for the study of Late- 
Celtic art in this country may be classified as 

follows : — 

Arms of Offence and Defence. 
Chariot Fittings. 
Personal Ornaments. 
Toilet Appliances. 
Domestic Appliances. 
Musical Instruments. 
Objects for Religious Use. 
Objects of Unknown Use. 

Pottery and Glass. 
Sepulchral Urns. 
Vessels for Domestic Use. 
Beads for Necklaces. 

Woodwork and Bonework. 
Vessels for Domestic Use. 


Sculptured Monuments. 




The arms of offence and defence of the Late-Celtic 
period are made of metal ; the sword-blades, dagger- 
blades, and lance-heads being of iron ; and the sword- 
sheaths, dagger-sheaths, shields, and helmets of bronze. 
In this country the bronze objects only are ornamented.^ 

Bronze sword- and dagger-sheaths have been found 
in considerable numbers in England, and also less 
frequently in Scotland and Ireland, as will be seen 
from the lists given below. 

List of Localities in Great Britain 'where Brofise Sword-sheaths 
of the Late-Celtic Period have been found. 



Houghton le Skerne 










Water Eaton 






Hod Hill 

Moreton Hall 


Bargany House 




Co. Durham. 

Co. Durham. 



















Co. Antrim. 

^ A lance-head of iron from La Tene, in Switzerland, is ornamented 
with engraved patterns, but nothing" of a similar kind has been found in 
Great Britain (E. Vouga, Les Helvetes a La Tene, pi. 5). 



List of Localities in Great Britain where Bro7ize Dagger- 
sheaths of the Late-Celtic Period have heeii found. 

River Witham . . . Lincolnshire. 

North Hinksey 
Southwark . 

Co. Galway. 

Some of these sheaths are elaborately ornamented, 
more especially the specimens from Hunsbury, Lisna- 
croghera, and the River Witham. The shape of the 
sheaths was evidently derived from a foreign source, 
as may be seen by comparing those found in Great 
Britain with the examples from Hallstatt and La Tene. 

Bronze shields of the Late-Celtic period are not by 
any means common, but the British Museum is fortunate 
enough to possess the only two perfect specimens now 
in existence. One of these came out of the River 
Thames at Battersea, and the other from the River 
Witham, near Lincoln. The former is, perhaps, on the 
whole, the most beautiful piece of Late-Celtic metal- 
work that has survived to the present time. It is of 
oblong shape with rounded corners like the Gaulish 
shields,^ and is made out of plates of thin hammered 
bronze, strengthened all round the edge by a roll 
moulding. The body of the shield consists of a plain 
plate upon which are riveted three circular pieces of 
ornamental repousse work, the largest one in the centre, 
and the other two smaller ones at the top and bottom. 
In the middle of each of the circular pieces of ornament 
is a raised boss, the annular space surrounding which 

^ See article on the Gaulish statue from Montdrag"on (Vaucluse) now 
in the Mus^e Calvet at Avig-non in the Revue Archdohgiqtie, N.S. , vol. xvi. 
(1867), p. 69; also Diodofus, bk. 5, ch. 30. 


k> m ^ \ 


Reproduced froDi k'anbic's " Hora Ferales" 


is filled in with gracefully flowing S- and C-shaped 
curves raised above the rest of the surface, and starting 
from and returning to small circular plaques of enamel 
with a swastika design on each. No written description 
can give any idea of the subtle decorative effect produced 
by the play of light on the surfaces of the flamboyant 
curves as they alternately expand and contract in width 
and rise and fall above the surrounding level back- 
ground. The drawing of the curves is simply exquisite, 
and their beauty is greatly enhanced by the sharp line 
used in all cases to emphasise the highest part of the 
ridge. It will be observed that the design is set out 
with regard to small circular bits of enamel placed in 
definite positions symmetrically round a central boss. 
If closely coiled spirals like those of the Bronze Age 
were to be substituted for the enamelled discs, we 
should then have a style of decoration exactly similar 
to that of the Christian Celtic MSS. The metalworker 
who made this shield seems to have possessed the true 
artistic feeling which told him instinctively exactly how 
much plain surface of shining bronze should be left to 
set off the ornament to the greatest advantage. The 
other shield in the British Museum, from the River 
Witham, is very inferior to the one just described, and 
is probably of later date. 

Late-Celtic bronze helmets are of great rarity. There 
are two in the British Museum, one from the Thames 
at London, and the other from an unknown locality. 
A third from Torrs, Kirkcudbrightshire, is now pre- 
served at Abbotsford, near Melrose. The specimen 
from the Thames is furnished with two conical horns 
terminating in small turned knobs, all the different 
pieces of wrought metal being riveted together with 
extreme neatness. The front of the helmet is orna- 



merited with small, round enamelled discs and repousse 
work in very low relief. The other helmet in the 
British Museum is shaped like a jockey's cap, and is 
particularly ugly in appearance. 

The helmet at Abbotsford has been so fully described 
by Dr. Joseph Anderson in his Scotland in Pagan 
Times : The Iron Age (p. 113) that it will not be neces- 
sary to say more about it here. 

Decorated bronze helmets of the La Tene period 
have been found in France at Berru^ (Marne) and 
Gorge-Meillet" (Marne). 

It will be seen from the list given below how extremely 
common finds of Late-Celtic horse-trappings have been. 

List of Localities in Great Britain ivhere Late-Celtic 
Horse-trappings have been found. 

South Shields 

. Co. Durham. 



Arras . 

. Yorkshire. 



Danes' Graves 


Kirkby Thore 




Locality unknown 




The Fens 


Saham Toney 


Westhall . 






Canterbury . 








Chessell Down 

Isle of Wight. 

^ A. Bertrand's Arch^ologie Celtique et Ganloise, 2nd ed. , 1889, p. 356. 
- E. Fourdrig^nier's Double Sepulture Gauloise de la Gorge-Meillet. 





Hag-bourn Hill 


Polden Hill . 


Hamdon Hill 

. Somersetshire. 



Neath . 


Clova . 





. Perthshire. 




















Co. Mayo. 

Clooncunra . 

Co. Roscommon. 


Co. Roscommon. 


Co. Meath. 


King's Co. 


Co. Monagfhan. 

Under the head of horse-trappings are included a 
large number of miscellaneous objects, such as bridle- 
bits, harness-rings, -buckles, and -mountings, pendants, 
head ornaments, etc. In fact, the term has been much 
abused by museum curators, who, when in doubt, say 
horse-trappings. Much the most important finds, con- 
sisting in each case of a large number of objects, have 
been those made at Polden Hill, Somersetshire, in 1801 ; 
Hagbourne Hill, Berks, in 1803; Westhall, Suffolk; 
Stanwick, Arras, and Rise, Yorkshire ; all the objects 
being now in the British Museum. The specimens 
from the Saham Toney find, which was equally im- 
portant, are to be seen in the Norwich Museum. 


Other smaller finds are preserved in the museums 
at Edinburgh and Dublin. 

Nearly all the big finds of horse-trappings have 
included several bridle-bits. These are usually quite 
plain, but there are, at least, four highly ornamented 
examples known (i) from Rise,^ Yorkshire, now in the 
British Museum ; (2) from Birrenswark,'^ Dumfries- 
shire, in the Edinburgh Museum ; (3) found near 
Tara,"' Co. Meath, now in the Dublin Museum ; and 
(4) from Kilkeeran,'^ Co. Monaghan, also at Dublin. 
These bridle-bits are formed of three or four separate 
pieces linked together, as in a modern one, and the 
decoration, which is concentrated on the terminal rings, 
consists of the usual Late-Celtic trumpet-shaped ex- 
pansions and coloured champleve enamels. 

In nearly all the finds of horse-trappings rings of 
various shapes and sizes are of frequent occurrence. 
They were probably used for passing the reins or other 
parts of the harness through, and perhaps also to act 
as strap buckles. Most of the rings are round in 
cross-section, except a segment separated from the rest 
by projecting flanges, the cross-section of which is 
made rectangular, apparently to enable the ring to be 
more rigidly fixed to the harness. The decoration of 
the rings usually consists of curious projections of 
various shapes, some resembling pairs of mushrooms 
placed with the convex tops together and the stems 
inclined at an angle ; whilst others are more like 
segments of an orange. Many of the rings are 
ornamented with engraved patterns composed of lines 

^ Magazine of Art for 1885, [). 456, 

- Dr. J. Anderson's Scotland in Pagan Times : Iron Age, p. 124. 

•5 Sir W. Wilde's Catal. of Mus, R. I. A., p. 605. 

^ Journ. R, Hist, and Archceol. Assoc, (f Ireland, N.S., vol. i., p. 423. 

Lower ends of Bronze Sword-sheaths from Hunsbury 

Now in the Northampton Museum 



and dots, or are enamelled. The best specimens in 
the British Museum have been derived from the finds 
already described at Stanwick, Yorkshire ; Polden 
Hill, Somerset ; and Westhall, vSuffolk. 

The harness-mountings are either in the form of a 
cross or a sort of rosette, with petals like a flower, 
some pointed and some round. At the back of the 
mounting are a pair of rectangular loops for passing 
a strap through. The front is, in many cases, beauti- 
fully enamelled. There is an extremely pretty little 
cruciform mounting of this kind in the British Museum, 
but unfortunately the locality whence it came is un- 
known. Two similar specimens have been recorded, 
one in the Uffizi Museum at Florence,^ and another from 
Saham Toney,' Norfolk, now in the Norwich Museum. 
The most elaborately decorated harness-mounting of 
the rosette type is the one from Polden Hill,'^ Somer- 
setshire, in the British Museum. 

A large number of objects found in Ireland, resem- 
bling a spur or the merry-bone of a chicken in shape, 
have been conjectured to be horses' head ornaments."* 
One of them was found near Tara, Co. Meath, with 
the bridle-bit already mentioned. 

Iron tyres of chariot -wheels have been found at 
Stanwick, Arras, Beverley, and Danes' Graves in 
Yorkshire, and Hunsbury, Northamptonshire ; but the 
bronze objects associated with them, which are believed 

^ Kemble's Horce Ferales, pi. 19, fig-. 5. 

2 Norfolk ArchcBology, vol. ii. , p. 398. 

^ Kemble's Hares Ferales, pi. 19, fig-. 3. 

■^ There are more than thirty-two in the Museum of the Royal Irish 
Academy (see Wilde's Catal., p. 109). Others have been found in the 
counties of Roscommon, Slig-o, and Cork (see Proc. R. I. A., vol. vii., 
p. 161 ; Vallancey's Coll. de Rebus Hibernicis, vol. iv. , p. 54 ; and Wood 
Martin's Pagan Ireland, p. 462). 










to be the fittings of the chariots, do not afford suffi- 
ciently characteristic decoration to need description 

The personal ornaments of the Late-Celtic period 
consist chiefly of fibulas, pins, collars, and armlets, 
usually of bronze, but in rare instances of gold or 

The evolution of the Roman Provincial type of 
fibula from earlier La Tene type can be nowhere better 
studied than in this country during the transition from 
the Late-Celtic to the Romano-British period. 

To anyone who is acquainted with the elaborate 
studies^ made by Scandinavian archaeologists on the 
origin and development of the various forms of fibulae 
found in northern Europe it must be a matter of surprise 
that up to the present no attempt has been made to do 
the same thing for our own country. With the ex- 
ception of Dr. Arthur Evans' paper in the Archceologia^- 
absolutely nothing has been written on the subject in 
England, nor do the curators of our public museums 
make the faintest attempt to classify the different kinds 
of fibulae of the Romano-British period according to 
their shapes. 

Looked at from a purely mechanical point of view, 
a fibula, or brooch, belongs to the same class of 
appliances as an ordinary door-lock ; being, in fact, a 
device for fastening applied to dress. The fibula was 
probably in its earlier stages evolved from a simple pin 
by endeavouring to invent some way by which the pin 
might be prevented from slipping out once it had been 

^ Hans Hildebrand's Industrial Arts of Scandinavia ; Oscar Montelius' 
" Spannen fran bronsaldern " in the Antiquarisk Tidskrift for Sverige ; 
and O. Almagren's Studien iiber norden europdische Fihelformen. 

^ Vol. Iv., p. 179. 


inserted in the fabric of the dress. A sufficiently 
obvious plan for effecting this is to connect the head of 
the pin with the point by means of a rigid bar sufficiently 
bent into the shape of an arch to avoid pressing too 
closely upon the portion of the dress between it and 
the pin. When fixed in its place the brooch forms a 
complete ring, so that a locking and unlocking con- 
trivance is necessary in order to enable it to be removed 
when not in use. 

The modern safety-pin, which is also one of the 
most ancient inventions, is perhaps the simplest kind 
of dress-fastener, and yet it is the parent of the almost 
endless series of European fibulag from the Bronze Age 
to the present time. It can be constructed in the easiest 
possible manner out of a single piece of metal wire of 
uniform thickness by making a coil in the middle of its 
length to act as a spring and a point at one end and a 
hook at the other. The pointed end is then bent round 
until it catches in the hook, and the thing is complete. 

There are two other classes of brooches which do 
not belong to the safety-pin type or its descendants, 
namely, (i) the Celtic penannular brooch ;^ and (2) the 
Northern Bronze Age brooch,^ which has a pin with 
a hole through the head enabling it to slide, turn, and 
move about loosely on the body of the brooch. With 
these we are not concerned at present. 

Although the safety-pin type of fibula was, in its 
earlier stages, made out of a single piece of wire, it 
may be considered to consist of four different parts, 
each of which performs a function of its own, namely, 
(i) the head, containing the spring or hinge; (2) the 
tail, containing the catch, or locking apparatus ; (3) 

^ Dr. J. Anderson's Scotland in Early Christian Times, 2nd ser., p. 7. 
2 J. J. A. Worsaae's Industrial Arts of Denmark, p. 92. 
















the body or framework, connecting the head with the 
tail ; and (4) the pin, moving on a hinge or spring at 
one end and with the pointed end fitting into the catch. 
In all fibulag derived from the safety-pin the pin is 
straight and the body bent into a more or less arched 
shape, like a bow. An infinite variety of forms were 
produced (i) by increasing the number of coils in the 
spring and their size ; (2) by expanding the tail end 
into a thin triangular plate ; and (3) by increasing the 
thickness of the body, or by making a coil in the 
middle of its length to act as a secondary spring. 
Much the most important modifications, however, in 
the safety-pin brooches were those which gradually led 
up to the harp-shaped, T-shaped, and cruciform fibula 
of the Romano-British period. Dr. Arthur J. Evans, 
in his paper in the Archceologia (vol. Iv., p. 179) on 
^'Two Fibulae of Celtic Fabric from vesica," has traced 
the evolution of the harp-shaped fibula from the bow- 
shaped fibula in a most interesting way. The different 
stages in the process appear to have been as follows: — 

(i) The tail end of the fibula was extended and bent 
backwards so as to make an S-shaped curve with the 
bow ; (2) the retroflected end of the tail was fixed to 
the middle of the convex side of the bow by means of 
a small collar, made in a separate piece ; (3) the whole 
of the back was formed out of one piece of metal, with 
the collar surviving as a mere ornament ; and (4) the 
triangular opening at the tail, bounded by the retro- 
flected end, part of the bow, and the catch for the point 
of the pin, was filled in solid with a thin plate. It will 
be noticed that during this process of evolution the 
extended and retroflected end of the tail has become 
part of the continuous curve of the convex side of the 
bow, whilst what was previously one half of the outside 


of the bow is now on the inside of the triangular plate 
at the tail end. This, together with the expansion of 
the head to suit the increased number of coils in the 
spring, produced the characteristic harp-shape of the 
Romano-British fibula, in many of which the knob 
ornament in the middle of the back is the last survival 
of the collar for fixing the retroflected end of the tail 
in its place. 

The cruciform and T-shaped fibulae, which began in 
Roman times and continued to be used by the Anglo- 
Saxons, resulted from extending the coils of the spring 
at the head symmetrically on both sides of the pin. 
In this class of fibula the two outside ends of the coil 
were joined by a loop passing through the inside of 
the bow so as to give extra leverage to the spring, or 
sometimes serving merely as a loop for suspension by 
means of a chain. 

A specimen of silver was found at the Warren,^ near 
Folkestone-, and is now in the British Museum. The 
lower portion is, unfortunately, broken off, but the 
retroflected end of the tail remains, with the little 
ornamental knob which is the survival of the practically 
useful collar for securing it to the back of the bow. 
The coils of the spring on each side of the pin and the 
connecting loop are clearly seen, together with the 
loose ring passing through the coils of the spring and 
a portion of the chain for suspension. 

An exceedingly pretty pair of harp-shaped fibulse of 
silver, with a well-wrought chain for suspension, were 
found near Chorley, Lancashire, with Roman coins 
dating from Galba to Hadrian, and are now in the 
British Museum. At the top of each fibula is a loop 
for attachment to the chain, and the bodies are beauti- 

^ The Reliquary for 1901, p. 197. 









mmt^ ii. WWiHli 






fully ornamented with Late-Celtic flamboyant patterns. 
The knob, which is the survival of the collar already 
referred to, has here assumed a highly ornamental form 
resembling two floriated capitals of columns placed 

The specimen represented on p. 104 is one of a pair 
of silver-gilt fibulas, similar to the preceding, but larger 
and without the chain, although possessing the loops for 
suspension. They were purchased in Newcastle about 
the year 181 1, and are now in the British Museum. It 
is stated in Hodgson's History of Northumherland 
(vol. iii., Appendix x., p. 440) that the locality from 
whence they came was somewhere in the county north- 
east of Backworth. The fibulas were discovered in a 
silver patera bearing a dedicatory inscription to the 
Dese Matres, and containing, in addition — 

5 gold rings. 

1 silver ring. 

2 gold chains with wheel pendants. 
I gold bracelet. 

3 silver spoons. 

1 mirror. 
280 denarii. 

2 large brass coins of Antoninus Pius. 

A full account of the find is given in E. Hawkins' 
'' Notice of a remarkable collection of ornaments of the 
Roman period, connected with the worship of the Deas 
Matres, and recently purchased for the British Museum" 
in the Archceological Journal {yo\, viii., p. 35). 

We may here call attention to the intensely Celtic 
character of the fibulae just described. The wearing of 
brooches in pairs with a chain attachment was a Celtic 
and not a Roman custom, as has already been pointed 



out in a previous volume of The Reliquary (for 1895, 
p. 157). A pair of bronze fibulse, of the same kind as the 
one from the Warren, Folkestone, fastened together 
by a double chain, was found in one of the Gaulish 
cemeteries in the Department of Marne^ in France, and 

One of a pair of silver-gilt Fibulaj found in Northumberland, 

with Denarius of Antoninus Pius (a.d. 139) 

Drawn by C. Piaetorius 

is now to be seen in the museum of St. Germain, near 
Paris. It may, therefore, be fairly assumed that all the 
fibulas found in this country with chains attached to 

^ Engraved in the Dictionnaire Archeologique de la Gaule. Other 
examples from the cemeteries of Somme Bionne, Courtois, Bussy-le- 
Chateau, and Sommesous in the Department of the Marne, are given in 
the Album accompanying L. Morel's La Champagne souterraine (pis. 13, 
29, 34, and 40). 


them or with loops for a chain at the top are more Celtic 
than Roman. 

Amongst the Late-Celtic antiquities in the British 
Museum are three specimens which illustrate the 
evolution of the harp-shaped fibula very well. One 
ornamented with a coral boss and gold stud, probably 
from the Marne district, was presented by the late Sir 
A. W. Franks ; another came from a chalk pit near 
Walmer, Kent ; and the third was found at Clogher, 
Co. Tyrone. 

Broadly speaking, it may be said that the safety-pin 
type of fibula made in one piece is earlier in date than 
the Roman occupation of Britain, and the specimens 
found in this country are obviously either imported 
from abroad or copied from foreign originals, such as 
those found at La Tene, in Switzerland, and in the 
Champagne district of France. The fibula in use in 
Britain, after it became a province of the Roman Em- 
pire, has a massive harp or bow-shaped back made in a 
separate piece from the pin and spring. In the earlier, 
or La Tene type of the fibula, the catch for the end of 
the pin forms one side of a triangular opening, which, 
as we have already mentioned, is filled in with a thin 
plate in the later or Roman Provincial fibula. There 
is also a sort of transitional kind, with ornamental pierc- 
ings in the plate. 

There was yet another description of fibula belonging 
to the Romano-British period, having a flat plate for 
the body in the shape of a circular disc, or sometimes 
in the shape of a fish or animal. 

The different classes of Late-Celtic fibulas are given 
in the following lists. 



List of Localities in Great Britain where Late-Celtic 
FibulcB have been found. 


Cowlam (Brit. Mus.) .... Yorkshire. 

Hammersmith (Brit. Mus.) 
Avebury (Brit. Mus.) . 
Water Eaton (Brit. Mus.) . 
Clogher (Brit. Mus.) . 

Co, Tyrone. 



Aylesford (Brit. Mus.) 

Folkestone (Brit. Mus.) 

Walmer (Brit. Mus.) . 

Locality not given (Liverpool Mus.) 




Ringham Low . 
Hod Hill (Brit. Mus.) 
London (Guildhall Mus.) . 
Bonville (Brit. Mus.) . 
Navan Rath (Mus. R.LA.) 
Locality unknown (Mus. R.LA.) 
Hunsbury (Northampton Mus.) . 





Co. Armagh. 

Co. Armagh. 




. Gloucestershire. 

Birdlip (Gloucester Mus.) 
London (Guildhall Mus.) 



Polden Hill (Brit. Mus.) . . . Somersetshire. 

Stamford Hill, Plymouth . . . Devonshire. 

Cricklade (Brit, Mus.) . . . Wiltshire. 





Backworth (Brit. Mus.) 

Chorley (Brit. Mus.) . 

Great Chesters (Newcastle Mus 

River Tyne (Newcastle Mus.) 

Rising-ham (Newcastle Mus.) 


Farley Heath 










Brough (Brit. Mus.) . 

Victoria Cave, Settle 

Silchester (Strathfieldsaye House) 





Kirkby Thore 

Dowkerbottom Cave, Settle 

Malton .... 

Thirst House Cave, Deepdale 

Kilnsea .... 


Locality unknown (Brit. Mus.) 







Metal pins do not seem to have been much used as 
dress-fasteners during the Late-Celtic period, judging 
from the number to be seen in our public museums. 
One of the most beautiful pins of this period now in 
existence is the one found with the burial previously 
mentioned at Danes' Graves,^ near Driffield, Yorkshire, 
and now in the York Museum. The pin is of bronze, 
with a peculiar crook near the top and a circular head 

^ Proc. Soc. Aiit. Lond. ser. 2, vol. xvii. , p. 120. 


(resembling a chariot-wheel with four spokes) inlaid 
with shell, or, according to another account, ena- 
melled. Two bronze pins, with plain turned heads, 
were amongst the objects derived from the Thirst 
House Cave,^ Deepdale, Derbyshire. 

Several pins of the class known as '' hammer-headed " 
have been discovered from time to time, chiefly in 
Ireland and Scotland. These pins are of considerable 
size, some being ten inches long, and have semi- 
circular heads with the convex side facing downwards. 
The top of the pin is bent at right angles, and the 
head fixed on in front of it. At the top of the head 
are usually from three to five projecting studs, and the 
face of the head is enamelled with Late-Celtic designs. 
From the associations in which such pins have been 
found and the style of their decoration, they would 
seem to belong to the transition period between 
Paganism and Christianity. There is one in the 
British Museum from Moresby, Cumberland, which 
was associated with a small bronze ornament of 
Late-Celtic character ; another in the same collection 
from Craigywarren,'- Co. Antrim, has spiral patterns 
upon it ; whilst a third in the Edinburgh Museum, 
from Norrie's Law," Forfarshire, was associated with 
coins of the seventh century, and silver leaf-shape pen- 
dants engraved with the same mysterious symbols 
which occur so frequently on the early Christian sculp- 
tured stones of Scotland. A hammer-headed pin of 
silver from Gaulcross,^ Banffshire, has spiral designs 
upon the head, but of a kind more nearly resembling 

^ The Reliquary for 1897, p. 96. 

- Wood Martin's Lake D-ivellings of Ireland, p. no. 

- Dr. J. Anderson's Scollatid hi Early Christian Times, 2nd sen, 
p. 36. 

^ Dr. J. Stuart's Sculptured Stones of Scotland, vol, ii. , pi. 9. 





.V Victor U'lnte, of Rcad)ii^\pJio:o. • :, 



that found on the Christian crosses of about the ninth 
century in Argyllshire than the Late-Celtic flamboyant 
designs of Pagan times. Other examples of pins of 
this kind have been found at Lagore,'^ Co. Meath, 
Urquhart,^ Elginshire, on the Culbin Sands, Nairn- 
shire, and in the island of Pabbay, Hebrides. 

Unquestionably the finest Late-Celtic personal orna- 
ments are the collars for wearing round the neck, of 
which, at least, two in gold and about ten in bronze are 
known to exist. Being larger than any other class of 
personal ornament, they naturally afford greater scope 
for the display of the elaborate forms of flamboyant 
designs in which the art metalworker of the period 
used to revel. One of the gold collars just referred to 
came from Broighter, on the western shore of Lough 
Foyle, near Limavady, Co. Londonderry. It was in 
the British Museum, but has recently been removed 
to Dublin. The collar, which formed part of one of 
the most valuable finds of gold ornaments yet made 
in Great Britain, is unique both as regards its form 
and the extraordinary artistic skill displayed in its 
decoration. The hoard was accidentally brought to 
light in 1896 whilst ploughing a field on the farm 
occupied by Mr. J. L. Gibson. We give a list of the 
various objects comprising the find below. 

List of Objects in the Limavady Find of Gold Ornainents. 

(i) Model of a boat, 7J inches long by 3 inches wide, 
weighing 3 ozs. 5 dwts., with benches and rowlocks for 
eighteen oarsmen (nine on each side) and rowlock for steer- 
ing-paddle in the stern. 

(2) Boat-fittings in miniature, consisting of fifteen oars, 

^ Wood Martin's Lake Dwellings of Ireland. 
^ Proc. Soc. Ant. Scot., vol. xxxv. , p. 279. 


one grappling-iron, three forked implements, one yard-arm, 
and one small spar, 

(3) Bowl, 3^ inches in diameter by 2 inches deep, weigh- 
ing* I oz. 5 dwts. 12 grs., provided with four small rings for 

(4) Two twisted necklets (one broken), the perfect one 
5 inches in diameter, weighing 3 ozs. 7 dwts. 9 grs. 

(5) Two chains of plaited wire, one i foot 2J inches long, 
weighing 2 ozs. 7 dwts., and the other i foot 4 J inches long, 
weighing 6 dwts. 12 grs. 

(6) Late-Celtic collar, yh inches in diameter, made of a 
tubular ring 1} inch in diameter. 

The collar must have had a joint of some kind, which is 
now missing ; and the fastening is a most peculiar one, 
consisting of a T-shaped projection on the end, one- 
half of the tubular ring fitting into a slot in the end of 
the other half of the ring. The locking is effected by 
giving the slotted end a half turn after the T-shaped 
projection has been inserted. The whole of the exterior 
surface of the tube is decorated with long sweeping 
curves, narrow in the middle and with trumpet-shaped 
expansions at each end, combined with helixes 
resembling a snail-shell. The background is shaded 
with a sort of engine-turned pattern of fine lines drawn 
with a pair of compasses. This remarkable gold collar 
has been fully described and illustrated by Dr. Arthur 
Evans in the Archceologia (vol. Iv., p. 397), and the facts 
relating to its discovery are related in detail by 
Mr. R. Cochrane, f.s.a., in a paper in the Journal of 
the Royal Society of Antiquaries of Ireland (vol. xxxiii., 
p. 21 1). An account of the evidence given by Mr. C. H. 
Read, f.s.a., before the committee appointed to inquire 
into the respective rights of the British Museum and 
the Museum of the Royal Irish Academy at Dublin to 





the possession of the hoard of gold ornaments will be 
found in the report of the inquiry in the Blue Book 
issued in 1899. 

A second collar of gold now in the Museum of the 
Royal Irish Academy in Dublin, said to have come 
from Clonmacnois, King's Co., is illustrated in Sir 
W. Wilde's Catalogue of Antiquities of Gold in Museum 
R.I.A.y p. 47. It consists of a plain hollow ring 
5J inches in diameter with an ornamental bulb on 
each side, one of which seems to be made in imitation 
of one of the glass beads of the period. 

The Bristol Museum possesses a perfect flat-jointed 
bronze collar, of a different kind from any of those 
just described, from Wraxhall,^ Somerset, and a por- 
tion of another from Llandyssyl,' Cardiganshire. In 
the British Museum there are two similar collars, one 
from Trenoweth,^ Cornwall, and the other from the Isle 
of Portland,* Dorsetshire. The Edinburgh Museum 
has also an exceedingly good example from Stitchell,-^ 
Roxburghshire. All these collars are elaborately or- 
namented in the Late-Celtic style. The date of the 
collar from the Isle of Portland is approximately fixed 
by its having been associated with a dish of Samian 

The existence of other Late-Celtic collars has been 
recorded at Mowroad,^ near Rochdale, Lancashire ; 
Embsay,^ near Skipton, Yorkshire ; Perdeswell,^ 
Worcestershire ; Lochar Moss,^ Dumfriesshire ; and 

^ ArchcBologia Camhrensis, ser. 6, vol. i., p. 83. 2 ll,i(^^ 

^ Archceologia , vol. xvi., p. 127. ^ Ihid. , vol. liv., p. 496. 

^ Dr. J. Anderson's Scotland i?i Pagan Times. 
^ H. Fishwick's History of the Parish of Rochdale, p. 5. 
"^ Archceologia, vol. xxxi. , p. 517. ^ Ihid., vol. xxx. , p. 554. 

^ D. Wilson's Prehistoric Aruials of Scotland, vol. ii. , p. 141, and 
Archceologia, vol, xxxiii., p. 347. 



Hyndford Crannog,^ near Lanark. These five be- 
long to a special class of what are not inaptly called 
''beaded torques," because rather more than one-half 
the collar is composed of bronze beads of two different 
shapes, (one convex and the other concave) strung 

Bronze Beaded Torque from Lochar Moss, Dumfriesshire 
Now in the British Museum 

alternately on an iron rod of square cross-section, so 
as to prevent the beads from revolving. The remain- 
ing and smaller segment of the circle consists of a 
bronze tube of rectangular cross-section, ornamented 
on the exterior with a Late-Celtic flamboyant design. 

^ Proc, Soc- Art. Scot., vol, jcxxiii., p. 385. 




The Perdeswell collar is incomplete, and the part which 
remains is formed of twenty beads resembling vertebra 
strung on to an iron wire or bar, as in the case of the 
Lochar Moss collar. 

The last class of personal ornaments of the Late- 
Celtic period to be noticed are the armlets. The most 
remarkable of these are of the Scottish type, as it may 
fairly be called, only one specimen having been found 
outside Scotland.! ^"^g armlets of this type are very 
heavy and massive, and their general form appears to 
have been suggested by a coiled serpent ; as in the one 
from the Culbin Sands, Nairnshire, the ends of the 
coil terminate in actual serpents' heads. The armlets 
are usually found in pairs, and are highly ornamented 
with flamboyant work, and in some cases enamelled. 
Although they are of cast-bronze, the style of the 
decoration is evidently copied from the repousse 
designs of the wrought metalwork of the period. 
Dr. J. Anderson has devoted a considerable portion 
of his Rhind Lectures on Scotland in Pagan Times: 
Iron AgCy to the examination of the Scottish group of 
armlets, most of which are in the Edinburgh Museum. 
The following is a list of the known examples : — 

List of Localities where A rmlets of the Scottish Type 
have been fo ii ncL 

Culbin Sands, Nairnshire. 
Auchenbadie, Banffshire. 
Castle Newe, Aberdeenshire. 
Belhelvie, Aberdeenshire. 
Aboyne, Aberdeenshire. 
Pitalpin, Forfarshire. 
Grange of Conan, Forfarshire 
Pitkelloney, Perthshire. 

1 At Newry, Co. Down. 

Bunrannoch, Perthshire. 
Seafield Tower, Fifeshire. 
Stanhope, Peeblesshire. 
Plunton Castle, Kirkcud- 


Locality unknown. 
Newry, Co. Down. 


The armlet from Stanhope, Peeblesshire, was asso- 
ciated with a Romano-British saucepan, which suggests 
that this type belongs to the later part of the Celtic 
Pagan Iron Age. 

Bronze armlets of La Tene, or continental type, 
have been derived from the burial mounds at Cowlam 
and Arras, Yorkshire. The bronze armlet from the 
Stamford Hill Cemetery, near Plymouth, is jointed like 
the collars, and decorated with flamboyant work. 

A pair of penannular ring armlets of silver terminating 
in sepents' heads, which may possibly be Late-Celtic, 
was disposed of at the sale of the Bateman Collection 
from Lomberdale House, Derbyshire. They were 
found at Castlethorpe,^ Buckinghamshire, in 1827, in 
a small urn containing Roman silver and brass coins, 
none later than the reign of Verus (a.d. 161-169), and 
a massive silver ring set with a carnelian engraved with 
a figure of Bonus Eventus. A similar pair of base 
silver armlets were found near the Carlswark Cavern, ^ 
in Middleton Dale, Derbyshire. 

Three very elegant armlets of twisted and looped 
bronze wire were associated with a Late-Celtic burial 
outside Thirst House Cave,'^ Deepdale, Derbyshire. 
Armlets of the same make are illustrated in Liden- 
schmit's Alterthiimer^ (vol. ii., pt. 5, pi. 3). 

The Late-Celtic toilet accessories are of three kinds, 
namely, hand-mirrors, hair-combs, and chatelaines. 
The mirrors are of bronze and circular in shape, with 
an ornamental handle. The back, or unpolished face 
of the mirror, is in nearly all cases decorated with 
incised circles of different sizes, combined with curved 

^ The Reliquary, vol. xiii., p], i8; -AnCfiJo^ir. ; and Brit. ArchceoJ. Assoc, 
vol. il., p. 353. 

'^ Til : Reliquary , \o\. 1867, p. 113. '* J bid., 1897, p. loi. 


^ 5 , > 

3 , ^ 3 

3 > 

^ :> -" 3 5 


lines and a peculiar sort of background filled in with 
cross hatching. A list of mirrors such as those de- 
scribed is given below. 

List of Localities whei'e Late-Celtic Mirrors have been found. 

Warden (Bedford Mus.) . . . Bedfordshire. 
Stamford Hill, near Plymouth (Ply- 
mouth Mus.) .... Devonshire. 
Birdlip (Gloucester Mus.) . . Gloucestershire. 
Trelan Bahow (British Mus.) . . Cornwall. 
Balmaclellan (Edinburgh Mus.) . Kirkcudbrightshire. 
Locality unknown (Liverpool Mus.) 

Unornamented mirrors have been found with burials 
at Arras/ Yorkshire, and Gilton,^ Kent. 

The hair-combs are of bone, and will therefore be 
described subsequently when dealing with bonework. 

The chatelaines of the Late-Celtic period are pretty 
little objects of bronze, generally enamelled. At the 
top is a loop for suspension ; there is a little rod below, 
from which are hung tweezers, picks, files, etc. Speci- 
mens have been discovered in the Thirst House Cave,^ 
Deepdale, Derbyshire, and at Canterbury,^ and Craven 
Arms,^ Shropshire. 

The domestic utensils and cooking appliances of the 
Late-Celtic period include wooden tankards and buckets 
with bronze mountings, bronze bowls and saucepans, 
and iron fire-dogs. Some of the riveted caldrons 
possibly also belong to this period, but as they cannot 
be distinguished from those of the Bronze Age it will 
be unnecessary to describe them here. 

^ W. Greenwell's British Barrows, p. 454. 

^ B. Faussett's Inventoritivi Sepulchrale, p. 30. 

^ The Reliquary for 1897, p. 95. 

^ Proc. Soc. Ant. Lond. , ser. 2, vol. vi., p. 376. 

^ ArchceoJogia Canibrensis, ser. 5, vol. vi., p. 90. 


There is a very perfect wooden tankard mounted 
with bronze in the Mayer Museum, Liverpool, from 
Trawsfynydd,^ Merionethshire, having a handle orna- 
mented in the Late-Celtic style with flamboyant tracery, 
which might easily be mistaken for Gothic work of the 
fourteenth century were it not for the trumpet-shaped 
expansions which occur in the details. Handles of 
similar tankards have been found at Aylesford,- Kent ; 
Elveden,^ Essex ; Okstrow,* Orkney; and Carlingwark 
Loch,'* Kirkcudbrightshire. 

Late-Celtic wooden buckets with bronze mountings 
are of the greatest rarity, so much so that only two are 
known to exist, one from Aylesford,** Kent, in the 
British Museum, and the other from Marlborough,'' 
Wilts, in the Devizes Museum. They are both deco- 
rated with repousse designs representing men, animals, 
etc., treated much in the same way as on the Ancient 
British and Gaulish coins of the same period. 

Bronze bowls have been frequently found on Late- 
Celtic inhabited sites and with Late-Celtic burials. A 
quite plain but extremely well-made bronze bowl is to 
be seen in the British Museum side by side with the 
beaded torque from Lochar Moss,^ Dumfriesshire, which 
accompanied it. There is another plain bowl in the 
Gloucester Museum which was associated with the 
burial at Birdlip,-* Gloucestershire, already described. 
A bronze bowl ornamented with projecting bosses is 

^ Archceologia Camhrensis, ser. 5, vol. xiii., p, 212. 

2 Archceologia, vol. Hi., p. 44. 

^ Ibid., vol. Hi,, p. 45. 

** J. Anderson's Scotland in Pagan Times : h-on Age, p. 242. 

^ Proc. Soc. Ant. Scot., vol. vii. , p. 7. 

''' The Reliqtiary for 1897, p. 35. 

7 Sir R. Colt Hoare's Ancient Wilts. 

^ D. Wilson's Prehistoric Annals of Scotland, vol. i., p. 465, pi. 9. 

* Trans, of Bristol and GloncestersJiire Archceol. Soc, vol. v., p. 137. 


amongst the objects derived from the Glastonbury^ 
Marsh Village ; and a bowl in the Dublin Museum 
from Keshkerrigan,^ Co. Leitrim, has a very character- 
istic Late-Celtic handle in the form of a beast made up 
of flamboyant curves. A special type of bronze bowls 
with zoomorphic handles and enamelled decorations 
will be dealt with subsequently. 

Most of the saucepans in use during the Late-Celtic 
period were either imported from Italy and Gaul or 
were so nearly copied by local metalworkers as to be 
indistinguishable from the originals. None of these 
saucepans, as far as I am aware, have Celtic decoration 
upon them, although several are inscribed with Celtic 
names, and others are highly enamelled. Two speci- 
mens in the British Museum are of exceptional interest, 
one of bronze enamelled and inscribed with the name 
^^BODVOGENVS," from Prickwillow,'^ near Ely, Cam- 
bridgeshire, and the other of silver, with a highly 
ornamented inscribed handle, which was found at 
Backworth,^ Northumberland, with the pair of Kelto- 
Roman fibulae previously mentioned. The more elabo- 
rate saucepans were probably used in connection with 
religious ceremonies and not for cooking, as is borne 
out by the dedicatory inscriptions upon the handles 
ard the circumstances under which many of them have 
1 een found. A list has already been given of the sauce- 
pans associated with finds of Late-Celtic objects. 

The metalworkers of the Late-Celtic period were not 
only capable of executing some of the finest pieces of 
repousse bronze that the world has ever seen, but they 

^ Proc. of Soniersetshire ArchcEol. Soc, vol. xL, p. 149, 

'^ Reliquary for 1900, p. 247. 

^ ArchcEologia, vol. xxviii., p. 436. 

^ Archceol. Jour., vol. viii., p. 39. 


also excelled in producing works of art in wrought- 
iron of great merit. As an example of their skill in 
this direction we have the remarkable pair of fire-dogs 
from Capel Garmon, Denbighshire/ now in the 
possession of Colonel Wynne Finch of Pentre Voelas, 
near Bettws-y-coed. The fire-dogs consist of two 
upright bars, each surmounted by the head of a beast 
with horns, and standing on an arched foot, connected 
near the bottom by a horizontal bar on which to rest 
the logs of wood used for the fire. The uprights are 
ornamented on each side with thinner pieces of iron 
bent into undulations and scrolls, and fixed to the up- 
rights at intervals with rivets having large round heads. 

Each of the beasts' heads has a very curious sort 
of crest ornamented with a row of circular holes and 
round knobs. Other fire-dogs of the same kind, made 
of plain iron bars, and with horned beasts' heads on 
the top of the uprights (each horn terminating in a 
round knob), have been found at Mount Bures,^ Essex, 
Hay Hill,'^ near Cambridge, and Stamfordbury,* Bed- 
fordshire, associated with Romano-British burials. 

The only objects of the Late-Celtic period which may 
conjecturally have been used for religious purposes are 
the little bronze figures of animals from Hounslow,^ 
Middlesex, now in the British Museum. 

Under the head of musical instruments come the 
bone flutes from Thor's Cave, Staffordshire, and the 
magnificent bronze trumpet found in 1794 at Loughna- 
shade, Co. Armagh, and now in the Museum of the 
Royal Irish Academy at Dublin. Most of the trumpets 

^ Archceologia Cambrensis, sen 6, vol. i., p. 40. 

'^ C. Roach vSmith's Collectanea Antigua, vol. ii., p. 25. 

•^ Archceologia, vol. xix. , p. 57. 

■* Publications of Cambridge Ant. Soc. for 1845 

^ Proc. Soc. Ant. Lond. , ser. 2, vol. iii., p. 90. 


of this kind are of the Bronze Age, but the style of 
the decoration on the annular disc at the mouth of the 
one from Loughnashade shows clearly that it is of 
the Iron Age.^ 

Amongst the objects of unknown use of the Late- 

Late Celtic Bronze Spoon from 
Brickhill Lane, London 

Late Celtic Bronze Spoon from 
Crosby Ravens worth, Westmoreland 

Celtic period are certain so-called spoons, some peculiar 
disc-and-hook ornaments, and a few highly ornamented 
circular pieces of repousse bronze with a cup-shaped 
depression nearly in the centre. 

The spoon-like objects have been very fully dealt 

1 Sir W. Wilde's CataL, pp. 627 and 631. 



with in a paper by Mr. Albert Way in the Archceological 
Journal (voL xxvi., p. 52), and below is given a list of 
all the known specimens. 

List of Localities 7v/iere Spoon-like Objects imth Late-Celtic 

Decoration have been found. 
Crosby Ravens worth (British Mus.) . Westmoreland. 
London, Brickhill Lane (British Mus.) . Middlesex, 

Late Celtic Spoon. One of a pair from Weston, near Bath. 

Now in tlie Edinburgh Museum. Scale \ linear 


London, Thames (British Mus.) 
Weston, near Bath (Edinburgh Mus.) 
Llanfair (Edinburgh Mus.) 
Penbryn (Ashmolean Mus.) 
Locality unknown (Liverpool Mus.) 
Locality unknown (Dublin Mus.) . 
Walmer ...... 








The body of these objects is shaped like a very 
shallow spoon with a pointed end, and the handle 
(if such it may be called) is circular or nearly circular, 
in many cases with two little round ears or projections 
at each side. The so-called spoons are generally found 
in pairs, one spoon having a cruciform design in the 
middle of the bowl ; whilst its fellow has a small hole 
bored through the edge of the bowl. The handles of 
the spoons are always ornamented, sometimes on the 
front only, but more commonly on the back as well. 

There are specimens of the other Late-Celtic objects 
of unknown use — namely, the hook-and-disc ornaments^ 
and the circular pieces of repousse metalwork with a 
cup-shaped depression — in the British Museum'- and 
the Dublin Museum.^ 

No satisfactory explanation has been given of the 
use of certain wheel and triskele pendants of which 
examples have been found in Berkshire, Kingsholm, 
near Gloucester, Hunsbury, N. Hants, Seamill Fort, 
Ayrshire, and Treceiri, Carnarvonshire. 


The pottery of the Late-Celtic period differs from 
that of the Bronze Age in being turned on a wheel 
instead of being hand-made. The firing is also better 

^ The Reliquary for 1901, p. 56. 

'■^ Proc. Soc. Ant. Scot., vol. xix., p. 254. 

•^ Sir W. Wylde's Catal. Mus. K.I. A., p. 637. 


done, and the quality of the ware superior in every 
way. Since the discovery of the Aylesford cemetery 
in Kent, in 1886, it has been possible to differentiate 
Late-Celtic pottery from Romano-British by the pecu- 
liar forms of the vases. Dr. Arthur Evans has dealt 
with this subject pretty exhaustively in his paper in the 
Archceologia (vol. Hi., p. 315). 

The most characteristic of the Aylesford urns is tall, 
with a narrow base and wide mouth. The base is in 
the shape of a low truncated cone, the top of which 
is the narrowest part of the vase, and from this point 
it gradually gets wider until the top rim is nearly 
reached, when it contracts again slightly. The curve 
thus produced is of such extreme elegance as to at 
once suggest a classical origin. The exterior surface 
of some of these pots is plain, but in many cases it 
is divided into bands by horizontal projecting bead 
mouldings. Dr. A. Evans does not find much diffi- 
culty in showing that the peculiarities of form can be 
directly traced to the metal situlce from which the vases 
were copied. With regard to this, he says : — 

" In most cases these {i.e. the Aylesford) vases, which for 
elegance of form may almost vie with the ceramic products 
of Italy or Greece, are divided into zones by the small raised 
ridges or cordons described above, the zones themselves 
being, in turn, decorated with finely incised linear striations. 
This type of vase, beautiful as it is in itself, is still more 
interesting from the comparisons to which it inevitably leads 
us. No one familiar with the ceramic forms of an important 
group of North-Italian cemeteries, belonging, for the most 
part, to the fourth or fifth centuries before our era, and of 
which the whole series of objects so admirably excavated 
and arranged by Professor Prosdocimi at Este^ forms the 
most splendid illustration, can fail to be struck with the 
^ Notizie degli Scavi, etc., 1882, pp. 5-37. 

latp::-celtic pottery from hunsbury 
now in the northampton museum 



manifold points of resemblance presented by the urns before 
us with the most characteristic of the vase-types there 
represented. The contour of the type referred to, with its 
shoulders sometimes angular, sometimes abruptly rounded 
off, its inverted conical body divided into vertical zones by 
raised cordons, and tapering off to a pedestal below, can 
only be described as identical with that of some of the finest 
of the Aylesford specimens. The only perceptible difference 
is that, whereas the British urns are almost uniformly 
covered with a black or brown coating — the colouring 

Late Celtic urns from Shoebury, Essex 
Now in the Colchester Museum 

matter may have been supplied by pounded charcoal — zones 
of the Euganean cineraries are coloured alternately with 
bands of graphite and red ochre. Some of the earlier of 
the Este vases are, however, of plain dark brown biicchero^ 
and others, again, of later date, of an uniform red or grey. 
These North-Italian parallels have a still further value, 
inasmuch as they throw the clearest possible light on the 
actual genesis of this type. The cordoned vases of Este 
are, in fact, nothing more than copies in clay of certain 
forms of bronze situ Ice ; the commonest form of these, 
which is distributed through the whole of the geographical 
area where these vases are discovered, is zoned in the same 


way as the pots, the zones answering to an universal method 
of early metal industry, in accordance with which vessels 
were built up of bands of thin metal riveted together at the 
edges, each zone being often, in turn, defined by cordons or 
beads of metal. These cordons themselves in their more 
prominent form represent the wooden rings that surrounded 
and kept together the framework of wooden staves, to which 
in early times the metal plates themselves were riveted." 

Besides the pedestalled vases from Aylesford,^ made 
in imitation of the cordoned sitiilcF of bronze from the 
North-Italian region, there are others, perhaps derived 
from them, with elegantly formed bases. There are also 
vases without pedestals, and having somewhat globular 
bodies as well as bowl-shaped and saucer-shaped pots. 
Most of these are now in the British Museum. 

The following list gives the finds of pottery of a 
similar kind : — 

List of Localities where Finds of Late-Celtic Pottery of the 
Aylesford Type have been made. 

Kit's Coty House (Maidstone Mus.) . . Kent. 

Allington (Maidstone Mus.) . . . Kent. 

Northfleet Kent. 

Elveden ....... Essex. 

Shoebury Essex. 

Braintree ...... Essex. 

Locality unknown (Cambridge Mus.). 

Hitchin Herts. 

Aston Clinton (Aylesbury Mus.) . . Bucks. 

Abingdon (Ashmolean Mus.) . . . Berks. 

Whitechurch (Dorchester Mus.) . . Dorset. 

Weymouth (British Mus.) . . . Dorset. 

Another class of pottery is recognised to belong to 
the Late-Celtic period, not so much by the forms of 

1 A fine example of this type from Sandy, Beds, is illustrated in 
T. Fisher's Bedfordshire. 







the vases (because most of them are in a very frag- 
mentary condition) as by the patterns upon them, 
which consist of incised curved lines, circles, dots, and 
different kinds of cross-hatching and shading. A list 
of the finds is given below. 

List of Localities where Fi7ids of Late-Celtic Potter \\ orna- 

mejited with Lncised Lines, Circles, Dots, and Shading, 
have been made. 

Hunsbury (Northampton Mus.) . Northamptonshire. 

Mount Caburn (Pitt-Rivers Coll.) . Sussex. 
(British Mus.) 

Brighton (Brighton Mus.) . . Sussex. 

Highfield Pits, near Salisbury (Black- 
more Mus., Salisbury) . . . Wiltshire. 

Kent's Cavern, Torquay (British Mus.) Devonshire. 

Glastonbury Marsh Village (Glaston- 
bury Mus.) ..... Somersetshire. 

Kingsholm (Ashmolean Mus.) . . Gloucestershire. 

Yarnton (British Mus.) . . . Oxfordshire. 

Those who wish to compare the Late-Celtic pottery 
of Britain with Gaulish pottery of the same character 
may, with advantage, consult the Dictionnaire Archeo- 
logiqite de la Gaule^ and Paul du Chatellier's La Poterie 
aux epoqiies prehistoriqiies et gauloise en Armorique. 

As far as the available evidence goes, glass does not 
seem to have been used for any other purpose by the 
Late-Celtic people except the manufacture of personal 
ornaments, the most important of which were beads 
for necklaces. Some of the beads from Ireland and 
Scotland, specimens of which may be seen in the 
museums at Dublin and Edinburgh, are most artisti- 
cally fashioned from twisted rods of glass of variegated 
colour bent into peculiar shapes. They have been 
obtained from the Irish crannogs at Lagore, Co. Meath, 
and Lough Ravel. 


A bracelet of green glass, with a cable-like ornament 
in white and blue strands surrounding its outer surface, 
was found a few years ago in the crannog at Hyndford, 
Co. Lanark. 


Owing to the perishable nature of the material very 
few examples of carved woodwork of the Late-Celtic 
period are now in existence. Those which we do 
possess have been derived from the Glastonbury Marsh 
Village and from the crannog at Lochlee, Ayrshire. 
Mr. Arthur Bulleid, f.s.a., has illustrated three speci- 
mens in an article on "Some Decorated Woodwork 
from the Glastonbury Lake Village" in the Antiquary 
for April, 1895, p. 109. No. i was dug up from the 
peat at a depth of 6 feet 6 inches below the surface, 
near the south-east edge of the village. It is a rect- 
angular piece of wood dressed smooth all over, i foot 
7 inches long by 3I inches wide by \ inch thick, 
decorated on one side with a step-pattern shaded after 
the fashion of chequerwork, with a cross-hatching of 
diagonal lines. No. 2 is the stave of a small bucket, 
which, when complete, must have been 7 inches high 
by 5i inches in diameter, decorated with a lozenge 
pattern shaded with parallel straight lines. No. 3 is a 
portion of a tub 6 inches high by i foot in diameter, 
cut out of a solid piece of ash, and having its exterior 
surface decorated with flowing lines of extreme beauty, 
resembling scrolls of foliage converted into geometrical 
ornament by successive copying. Where the flowing 
lines diverge, the trumpet- shaped expansions are 
shaded with diagonal cross-hatching and dots. There 
is a good model of this tub in the British Museum. 
The designs on the woodwork from Glastonbury are 


produced by incising the surface with some fine sharp- 
pointed tool, and afterwards burnt in by passing a 
heated piece of metal along the incisions. 

The specimen from the Lochlee crannog, which is 
illustrated in Dr. R. Munro's Ancient Scottish Lake 
Dwellings (p. 134), is a piece of ash 5 inches square, 
ornamented on one side with a triple spiral, and on the 
other with Late-Celtic flamboyant work. 

A wooden bowl with a carved handle, found in a 
bog near Rathconrath, Co. Westmeath, and now in 
the Museum of the Royal Irish Academy in Dublin, 
may possibly belong to the same category. 

Amongst the objects of bone belonging to the Late- 
Celtic period the most remarkable are the spatulse, or 
flakes, of which no less than 5,000 are said to have been 
derived from cairn H of the Slieve-na-Caillighe series, 
near Oldcastle, Co. Meath. These chambered cairns 
were in the first instance erected as burial-places at the 
end of the Neolithic Age or the beginning of the 
Bronze Age, and the one marked H on the plan given 
in the Proc, Soc. Ant. Scot. (vol. xxvi., p. 294) appears 
to have been used as a workshop by an artificer in bone 
during the Early Iron Age. Ninety-one of the bone 
spatulse from the cairn in question were engraved by 
compass, with circles, curves, and ornamental punctur- 
ings, and twelve were decorated on both sides. Un- 
fortunately the whole of the bones have been lost, and 
we only know what they were like from the illustrations 
in E. Conwell's Tomb of Ollamh Fodhla (p. 53). A frag- 
ment of one of these bones which had been overlooked 
by the previous explorers of the cairn has recently been 
brought to light by Mr. E. Crofton Rotheram.^ Perhaps 
the most interesting feature connected with the bones 
from Slieve-na-Caillighe is the discovery with them of 

'^ Journ, R. Soc. Ant. Ireland, ser. 5, vol. vi., p. 257. 


the pair of iron compasses used in producing the in- 
cised designs upon them. 

Besides the bones just described, the other principal 
objects of the same material belonging to the Late-Celtic 
period are certain toilet-combs and spoon-shaped fibula, 
or dress-fasteners. Bone combs with Late-Celtic orna- 
ment have been found on the inhabited site at Ghegan 
Rock, near Seacliff, Haddingtonshire, and in the cran- 
nogs at Lagore,^ Co. Meath, Ballinderry,- Co. West- 
meath ; and at Longbank crannog on the Clyde, near 
Glasgow. Spoon-shaped fibulas of bone have been 
derived from the Victoria Cave, Settle, the Kelko Cave, 
Giggleswick, and Dowkerbottom Cave, Arncliffe, York- 
shire. The ornament upon them consists of concentric 
circles and dots.^^ 

In addition to wood and bone, the Late-Celtic people 
used Kimmeridge shale for the manufacture of objects, 
chiefly turned vases with cordons, like the Aylesford pots 
previously described. Vessels of this kind have been 
found at Old Warden,'* Bedfordshire, Great Chester- 
ford^ and Colchester,'^ Essex, and Corfe Castle,^ Dorset. 


Only three sculptured monuments decorated with 
Late-Celtic patterns are known to exist at present.^ 
They are all in Ireland and are fully described by 
Mr. G. Coffey in the Proceedings of the Royal Irish 
Academy (vol. xxiv., sect, c, p. 257). 

'' Sir W. Wilde's Catal. Mus. R.LA., p. 271, fig-. 176. 

^ Ibid., p. 271, fig. 177. 

" Prof. W. Boyd Dawkins' Cave Hunting, pp. 91 and 131. 

■* "On the Materials of Two Sepulchral Vessels found at W^arden, 
Co. Beds, by the Rev. J. S. Henslow {Cambridge Ant. Sor. PkU. , 1846). 

'' ArchcBol. Jotir. , xo\. xiv., p. 85. 

^ Henslow, loc. cit , p. 87. '' Archceol. Jour , vol. xxv., p. 301. 

^ At MuUag-hmast, Co. Kildare ; Castle Strange, Co. Roscommon ; 
and Turoe, Co. Galway. 



Ref7vdnced fi-oiii ct pliotcsrapli by Mr. -I. McGoogan 
iHustratiiiL; Mr. George Coffey's paper in the 
'' Proceedinrx 0/ the Royal Irish Aeadeviy'' 

) 3 :> 

^ 3 5 J 

^ 3 J 




THE fact must never be lost sight of that the 
picture presented to our mind of any particular 
prehistoric stage of culture must necessarily be 
extremely imperfect, since the extent of our knowledge 
is limited entirely by the number of relics which 
specially favourable circumstances have preserved from 
destruction. Of the textile fabrics of the Late-Celtic 
period, for instance, hardly anything is known, al- 
though we are certain that spinning and weaving 
must have been extensively practised from the quan- 
tities of long-handled weaving-combs, spindle-whorls, 
and loom-weights that have been found on almost 
every inhabited site. A people who showed such a 
high capacity for decorative design could not have 
failed to produce good artistic effects by means of 
pattern-weaving.^ What such textile patterns may have 
been can only be guessed at by survivals (like the 
Scottish tartans) or by ornament of a textile character 
occurring on objects made of less perishable materials 
(like the step-pattern on a piece of wood from the 
Glastonbury Marsh Village). 

^ "The cloth was covered with an infinite number of little squares 
and lines as if it had been sprinkled with flowers, or was striped with 
crossing- bars which formed a chequered design. Their favourite colour 
was red or a pretty crimson." C. Elton's Origins of English History, 
p. 114, quoting Pliny and Diodorus Siculus. 
K 129 


The Celts had already become expert workers in 
metal before the close of the Bronze Age ; they could 
make beautiful hollow castings for the chapes of their 
sword-sheaths ; they could beat out bronze into thin 
plates and rivet them together sufficiently well to form 
water-tight caldrons; they could ornament their circular 
bronze shields and golden diadems with repousse pat- 
terns, consisting of corrugations and rows of raised 
bosses ; and they were not unacquainted with the art 
of engraving on metal. 

The Celt of the Early Iron Age attained to a still 
higher proficiency in metallurgy than his predecessor 
of the Bronze Age. Casting in bronze was applied to 
a much larger number of objects than before, such as — 

Handles of swords and daggers. 

Chapes of sword and dagger sheaths. 


Harness-mountings and rings. 

Chariot fittings. 

Collars and armlets. 

Handles of tankards and mirrors. 

Spoon-like objects of unknown use. 

Wrought-bronze was used for — 
Sword and dagger sheaths. 
Shields and helmets. 

Mountings of wooden buckets and tankards. 
Caldrons and buckets. 
Circular discs of unknown use. 

The ornamental features of the objects of cast-bronze 
were produced chiefly during the process of moulding, 
although the surface was in many cases further beauti- 
fied afterwards by chasing, engraving, and enamelling. 

Objects of wrought-bronze were usually decorated 
by means of repousse-work, i.e. designs in relief 





Late-Celtic Bronze Mirror from Trelaii Bahow, Cornwall 

Now in the British Museum 


hammered up from the back. Occa- 
sionally enamel was added (as, for 
example, on the shield from the 
Thames, now in the British Museum), 
in the form of small plaques fixed 
on with rivets. In place of the more 
or less crude corrugations and rows 
of raised pellets of the Bronze Age 
we get the most marvellous curved 
surfaces and conchoids, executed 
with an unerring eye and a skill 
which it would be difficult to surpass. 
The repousse-work of the Late-Celtic 
period is seen in its greatest perfec- 
tion on the circular discs of unknown 
use^ in the Museum of the Royal 
Irish Academy in Dublin, and the 
British Museum. 

Both cast and wrought objects of 
bronze were decorated with patterns 
composed of finely engraved lines 
shaded in places with a peculiar kind 
of cross-hatching or with dots. The 
mirror- backs, ^ the sword -sheaths,^ 
and the harness-rings^ afford good 
examples of this class of work. 

Brazing and soldering appear to 
have been unknown to the metal- 
workers of the Late-Celtic period, as 

they pieced their metalwork together by means of rivets. 

The practice of riveting was learnt from the artificers 

Upper part of Bronze 

Sword-sheath from 


Now in the Northanpton 

^ See p. 12 1. "-^ See Hst on p. 115. 

^ See list on p. 91 ; especially the one from Lisnacrog'hera. 

■* See list on p. 94 ; especially those from Polden Hill, Somerset, 


who constructed the caldrons of the late Bronze Age 
already referred to, and they, no doubt, in their turn, 
acquired their knowledge from a foreign source. The 
bronze helmet from the Thames at Waterloo Bridge, 
now in the British Museum, illustrates the riveting of 
the Late-Celtic period at its best. The rivets generally 
have pointed conical heads, producing a good decorative 
effect. The way in which the different pieces of metal 
are held together is often ingeniously disguised by 
making the rivet-heads form part of the ornament, or 
by concealing the head behind a circular disc of enamel. 
The evidence of both history and archeology tends 
to show that the art of enamelling on metal was, in the 
first instance, a British one. The historical evidence 
is confined to an oft-quoted passage from the Icones of 
Philostratus (a Greek sophist in the court of Julia 
Domna, wife of the Emperor Severus), which is as 
follows : — 

^'They say that the barbarians who live in the ocean pour 
these colours on heated brass, and that they adhere, become 
hard as stone, and preserve the designs that are made upon 

Philostratus wrote this at the beginning of the third 
century a.d. ; and by "the barbarians who live in the 
ocean" (roi/? eV'f2/ceai/co ^ap^apov^) he no doubt meant 
the Britons rather than the Gauls, as some French 
writers have assumed. ^ 

The earliest enamels are those which occur on objects 
decorated in the pure Late- Celtic style without any 
trace of Roman influence, such as — 

Bridle-bits from Rise, near Hull, Yorkshire ; and Birren- 
swark, Dumfriesshire. 

^ A. W. Franks in Kemble's Horce Ferales, p. iS6. 



Harness-riiig-s and mo 

intings from 

Alfriston . 
Polden Hill 
Saham Tone} 
Uffizi Museum 
British Museum 
Armlets from Castle Newe, Aberdeenshire , 
loney, Perthshire. 

Handles of bowl from Barlaston, Staffordshire. 








Locality unknown. 

and Pitkel- 

Next, in order of age, come objects which from their 
general form or from the associations they were found 
in are known to belong to the Romano-British period, 
but yet have Late-Celtic decoration upon them, such 
as — 

Harp-shaped fibulae from Risingham, Northumberland, 
River Tyne, in the Newcastle Museum. 

S-shaped fibuke from Norton, Yorkshire, and other places 
(see list on p. 107). 

Seal-box from Lincoln, in the British Museum. 

Four-legged stand, with round hole in the top, from 
Silchester, in the Reading Museum. 

Lastly, there are survivals of the use of discs of 
Late-Celtic enamel in the decoration of bowls of early 
Saxon, and therefore post-Roman, age, the following 
examples of which have been found : — 

List of Localities 2v/ierc Bowls of the Saxon Period^ hut with 
Late-Celtic eiianielled decoration^ have been found. 
Crosthwaite (British Mus.) . . Cumberland. 

Middleton Moor (Sheffield Mus.) . Derbyshire. 
Over-Haddon .... Derbyshire. 

Benty Grange .... Derbyshire. 


Chesterton (Warwick Mus.) . 


Caistor (now lost) .... 


Oxford (Pitt-Rivers Collection) 


Needham Market (now lost) . 


Barring-ton (Sir John Evans' Col- 



LuUingstone (Sir W. Hart Dyke's 

Collection) .... 


Greenwich (Mr. J. Brent's Collec- 

tion) ..... 


Kingston Down .... 


The hammer-headed pins, a list of which has already 
been given on page io8, are also instances of the use 
of Celtic enamel in post-Roman times. 

Before going further it may be as well to say a 
few words about the art of enamelling in general, so 
as to show the position occupied by the Late-Celtic 

The term enamel is used to designate a particular 
kind of mixture or paste which can be applied to the 
surface of metals or other materials, so that when it has 
been vitrified by the application of heat, and afterwards 
cooled, it forms a decoration of great beauty and 
durability. The base of all enamels is a flux composed 
of silica (in the shape of silver sand or powdered flint), 
red lead, and potash. To this flux are added certain 
metallic oxides to produce different colours, and, if 
necessary, oxide of tin to render it opaque. The 
materials are mixed together, fused in a crucible, 
reduced to a fine powder when cold, made into a paste 
with water, and then applied to the surface of metal to 
be decorated. After vitrifaction in a furnace and polish- 
ing, the enamel is complete. 


Mr. A. W. Franks^ divides enamels into the following 
classes : — 

(i) Inlaid Enamel^ where the outlines are formed by metal 

(2) Transpare7it Enamel^ where the outlines and all the 
markings are produced by variations of depth in the sculp- 
tured ground over which the vitreous material is floated. 

(3) Painted Enamel^ where the outlines are made by a 
difference in the tint of the enamel itself, which completely 
conceals the metal base beneath. 

The divisions between each of the colours in inlaid 
enamel are produced in two different ways, namely : — 

(<?) Champleve Enamel, where the field (champ) or area 
to be occupied by each colour is dug out and removed ileve)^ 
so as to leave a very narrow band of metal at the level of 
the original surface to form the dividing line between the 

{h) Cloisonne Enamel^ where the divisions or partitions 
{cloison) between the fields consist of thin strips of metal 
bent into the required shape and fixed to the surface to be 

All the enamels of the Late-Celtic period belong to 
the champleve kind. The colours used are bright red, 
yellow and blue, and the designs are more often curvi- 
linear than not, like those on the repousse metalwork. 
The patterns were probably traced on the surface of 
the metal to be decorated with a finely pointed instru- 
ment, and the hollows to receive the enamel dug out 
with a scooping tool, in the case of small work, or with 
a long thin chisel and a chaser's hammer where the 
work was larger. 

The late Sir Wollaston Franks, than whom no better 

^ Afterwards Sir Wollaston Franks ; see Glass and Enamel, by J. B. 
Waring- and A. W. Franks. 






SCALE ^ linear' ^ ^i\ \i'' ' ^^' y 


authority can be quoted, always maintained the Celtic 
origin of the art of enamelling in Western Europe, 
and gave the distinctive name of opus Britanniciim to 
the special kind of enamel which was produced in 
greater perfection by the Celts inhabiting the British 
Isles than by any other people. The art of enamelling 
in the purely Celtic style commenced before the arrival 
of the Romans in this country, and after continuing 
throughout the whole period of their occupation, sur- 
vived for some centuries after their departure from our 
shores. There are, however, numerous enamels which, 
though very possibly of Celtic workmanship, are 
altogether Roman as far as the ornamental patterns 
upon them are concerned. Dr. Joseph Anderson has 
described an exquisitely enamelled patera of this kind 
found in Linlithgowshire, and now in the Museum of 
Antiquities of Edinburgh. He says of it : — ^ 

''Apart from the singular beauty of its decoration it is 
possessed of this special interest, that it is the only vessel of 
its kind and character known to exist in Scotland. It is, 
however, one of a class of objects, which, though few in 
number, are pretty widely distributed over the area, which 
may be termed the outskirts of the Roman Empire, towards 
the north and west — that is Britain, North Germany, and 
Scandinavia. We look in vain for anything like it within 
the area of the Roman Empire proper, and it may therefore 
be regarded as a product of a culture of some portion of the 
area of north-western Europe, where it was touched and 
modified by the Roman culture." 

Other similar examples of enamelled vessels have 
been found at Braughing,^ near Standon, Herts ; the 

^ " Notice of an enamelled cup or patera of bronze found in Linlith- 
g"owshire, recently purchased for the Museum," in the Proc. Ant. Scot., 
vol. xlx., p. 45. 

■^ Proc. Soc. A?it. Lorid., 2nd sen, vol. iv. , p. 514. 


Bartlow Hills/ Essex; Maltbeck,^ Denmark; and 
Pyrmont,^ in the Rhine valley. In addition to these 
we have two other enamelled vessels, but differing in 
their style of ornament, one from Rudge,* Wilts, now 
in the Duke of Northumberland's private museum at 
Alnwick Castle, and the other from Prickwillow,^ Cam- 
bridgeshire, now in the British Museum. 

Of the art of enamelling as carried on elsewhere than 
in Britain Dr. Anderson says : — ^ 

"The Gauls as well as the Britons — of the same Celtic 
stock — practised enamel-working before the Roman conquest. 
The enamel workshops of Bibracte, with their furnaces, 
crucibles, moulds, polishing -stones, and with the crude 
enamels in their various stages of preparation, have been 
recently excavated from the ruins of the city destroyed by 
Caesar and his legions. But the Bibracte enamels are the 
work of mere dabblers in the art compared with the British 
examples. The home of the art was Britain, and the style 
of the patterns as well as the associations in which the 
objects decorated with it were found, demonstrate with 
certainty that it had reached its highest stage of indigenous 
development before It came in contact with the Roman 

A full account of the discoveries made at Bibracte 
will be found in J. G. Bulliot's Fouilles de Mont 
Beuvray. Several beautiful enamels have been derived 
from the Belgo-Roman cemetery at Presles. 

Romano-British enamels, without distinctively Celtic 

^ ArchcEoIogia, vol. xxvi., p. 300. 

- Mem. de la Soc. Royale des Antiquaires du Nord, 1866-71, p. 151. 
•" Jahrbilcher des Vereins voii AltertJiionsfreiinden in Rheinlande, 
heft 38, p. 47. 

^ Catal. of Mns. of R. Archceol. Insf. at Edinburgh, 1856, 
^ Archceologia , vol. xxviii., p. 436. 
^ Loc. cit., p. 49. 



patterns upon them, have been dug up at many places 
in Great Britain, but more especially at Prickwillow. 

We shall see in a subsequent chapter how the diver- 
gent spiral patterns on the circular discs of enamel 
used to decorate the bronze bowls of the end of the 
Late-Celtic period were transferred bodily to the pages 
of the early Irish illuminated MSS. of the Gospels. 

Another method of ornamenting metalwork besides 
enamelling was by means of settings of different 
materials fixed in place by small pins or rivets. As 
instances we have the bronze shield from the River 
Witham,^ now in the British Museum, set with red 
coral ; the bronze fibula from Datchet, Oxon,- set with 
amber and blue glass; and most curious of all, a 
bronze object of unknown use from Carlton,^ North- 
amptonshire, now in the Northampton Museum, inlaid 
with portions of the stem of a fossil encrinite. 

A very effective kind of decorative metalwork may 
be made out of wire, bent so as to form a series of 
loops, of which we have British examples in the brace- 
lets from the Early Iron Age burial in Deepdale,^ 
Derbyshire ; and a foreign specimen in a fibula from 
the cemetery of the La Tene period at Jezerine,^ in 

More or less akin to the looped wirework just men- 
tioned are certain gold and silver chains made of fine 
wire. Dr. Arthur Evans has gone pretty fully into 
this subject in his paper in the Archceologia (vol. Iv., 
p. 394), describing the find of gold ornaments at 
Broighter, near Limavady, Co. Londonderry, amongst 

^ J. Kemble's Horce Ferales, p. 14, and pi. 15. 

- Proc. Soc\ Ant. Lond., ser. 2, vol. xv, , p. 191. 

"^ Ibid., ser. 2, vol. xvii., p 166. 

** The Reliqnary for 1S97, p. loi, 

^ R. Munro's Boznia-Herzegovina, p. 170. 


which were two chains of the kind referred to. The 
art of making these chains was no doubt of foreign 
origin, as they have been found in Etruscan tombs of 
the fifth century B.C. in Italy; with burials of the La 
Tene period in the cemetery of Jezerine, in Bosnia ; 
and in a tomb in the Gaulish cemetery of Ornovasso, 
in the province of Turin. In Britain such chains were 
used during the period of the Roman occupation for 
the attachment of fibulse worn in pairs, as in the case 
of those from Chorley,^ Lancashire, and from New- 
castle-on-Tyne,'- Northumberland. We shall see in a 
subsequent chapter that the manufacture of these finely 
wrought chains of silver survived in early Christian 
times in Ireland, the best-known examples being those 
attached to the Tara brooch,'^ and to an enamelled pin 
from Clonmacnois.^ With regard to the date of the 
chains. Dr. Arthur Evans says: — '* 

" It thus appears that these fine chains were in use among 
the Celtic peoples during the first two centuries before and 
after our era.'' In Britain, however, the finest class is, as 
far as I am aware, confined to the latter half of this period ; 
the chains attached to the earlier British fibulse, like the one 
In the British Museum from the Warren, *" near Folkestone, 
which may date from the second century B.C., being, like those 
referred to from the Champagne^ cemeteries, of simpler and 
coarser construction." 

^ The Reliquary for 1901, p. 198. 

- Ihid. for 1895, p. 157. 

^ M. Stokes' Early Christian Art in Ireland, p 75. 

^ Jour. R, Soc. Ant. of Ireland, ser. 5, vol. i., p. 318. 

■^ Archceologia , vol. Iv, , p. 396, 

^ Dr. A. Evans appears to have forgotten the Christian survivals in 

'^ The Religtmry ^ov 1901, p. 197. 

^ The coarser chains are made of ordinary circular or oval links, 
sometimes double (see illustrations given in the Dictionnaire Arch^ologique 
de la Gaule of those from the Marnian cemeteries). 



Ornamental ironwork of the Late-Celtic period is 
extremely rare, either because the smiths were too 
busily employed in making weapons for the warrior 
and tools for the artisan to devote their time to de- 
corative work, or because the specimens of their 
handiwork have disappeared in consequence of the 
perishable nature of the material of which they were 
made. Fortunately, however, the fire-dogs from Capel 
Garmon,^ now at Colonel Wynne Finch's house at 
Voelas, are still in existence to show us what fine 
ornamental ironwork the Welsh smiths of the Romano- 
British period were capable of producing. 

Turning now from the metalwork to the pottery of 
the Late-Celtic period, we find it to consist of unglazed 
vessels made on a wheel, fired in a kiln, and ornamented 
either by mouldings or by patterns engraved on the 
surface with a pointed instrument. The technical 
processes employed in its manufacture do not seem to 
have differed essentially from those of the Romano- 
British potters, except that slip-ware was unknown. 
As far as I am aware, no painted pottery like that from 
Mont Beuvray^ (Bibracte), nor vessels incrusted with 
pebbles and polished with graphite like the one from 
Plouhinec^ (Finistere) now in M. Paul du Chatellier's 
collection at the Chateau de Kernuz, near Quimper, 
have yet been discovered in this country. 

As far as the existing evidence goes no ornamental 
glasswork was made during the Late-Celtic period in 
Britain except certain beads and armlets already de- 
scribed. The technical process of manufacturing these 
beads consisted in twisting together fine rods of different 

^ ArchcEohgia Cambrensis, sen 6, vol. i. , p. 39. 
^ See J. G. Bulliot's Fouilles de Mont Beuvray. 
■"' See Revue ArcJi^ologique for 1883, p 11. 



coloured glass, and then bending the composite rod 
into loops round a mandril so as to form the bead. 

The art of the ornamental worker in wood in the 
Late-Celtic period is displayed at its best in the 
tankards, buckets, and tubs of which, fortunately, a 

1-75x25 JX?J 8-7/X6 

Liite-Celtic Pottery from the Glastonbury Lake Villag-e 

few interesting specimens have been preserved. The 
tankard from Trawsfynydd,^ Merionethshire, now in 
the Liverpool Museum, shows great ingenuity of con- 
struction, the staves of which it is composed being 
kept together at the bottom by a corrugated wire let 

^ Arch(£ohgia Camhrensis, ser. 5, vol. xiii., p. 212. 

'in the early iron age 143 

into the ends of the staves. Another tankard, belong- 
ing to Mr. T. Layton,^ f.s.a., has the staves fastened 
together with wooden dowels and pins. 

We have already described how the engraved patterns 
were produced on the ornamental woodwork from the 
Glastonbury Marsh Village by a finely pointed instru- 
ment, and afterwards burnt in. 

Ornamental objects were also made out of bone and 
Kimmeridge shale during the Late-Celtic period, but 
there is nothing special to call for any comment in the 
technical methods employed, except to mention that 
the patterns on the bone objects were often engraved 
by means of a pair of compasses, and that the vessels 
of Kimmeridge shale were turned on a lathe. 

Of the basketry in which the Celts excelled in the 
time of Caesar-^ no specimens are now extant, but no 
doubt their natural talent for decorative art showed 
itself in this native industry of Britain, as in all others. 


Once the eye of a trained archaeologist has become 
familiar with the general appearance of the art products 
of the Late-Celtic school, it is comparatively easy for 
him to recognise other products of the same school 
almost by intuition ; but he would find it a much more 
difficult task if he were asked to define exactly what the 
peculiarities are by which he is enabled to distinguish 
this particular style from any other. Most of the 
decorative elements composing the style are of so 
fantastic and original a nature as to impress themselves 
first on the retina of the eye, and then on the mind ; 
yet, on that very account, they seem to elude the 

^ Archceologia , vol. lii., p. 359. 

'^ C. Elton's Origins of Etiglish History, p. 122. 


descriptive powers of the writer. The motives em- 
ployed are neither purely geometrical in character nor 
have they been obviously arrived at by conventional- 
ising natural forms, but are something between the two, 
being (like the designs on the ancient British coins) the 
result of successive copying. We will, however, not- 
withstanding the difficulties that have been pointed out, 
endeavour to analyse the decoration of the Late-Celtic 
period as far as it is possible to do so. 

Unlike the art of the Bronze Age, the art of the Late- 
Celtic period does not appear to have been in any way 
influenced by religious symbolism, and therefore must 
be looked upon as purely decorative. The designs 
may be divided into three classes as regards the method 
of their execution, namely : — 

(i) Designs engraved on a flat surface. 

(2) Designs in relief on a flat surface. 

(3) Designs in the round. 

The designs themselves may be classified as follows : — 

(i) Anthropomorphic designs. 

(2) Zoomorphic designs. 

(3) Designs derived from foliage. 

(4) Curvilinear geometrical designs. 

(5) Rectilinear geometrical designs. 

Anthropomorphic and zoomorphic designs are ex- 
tremely rare in Late-Celtic art in Great Britain, and 
the two best-known examples — the buckets from Marl- 
borough^ and Aylesford"^ — have, according to Dr. Arthur 
Evans, been imported from Gaul. The Marlborough 
bucket is encircled by four horizontal metal bands, the 
upper three of which are decorated with human heads 

^ Sir R. C. Hocire's Ancient Wills, vol. ii, , p. 34, and W. Cunning-ton's 
Catal. of Stourhead Coll. in Devizes Museum, p. 88. 
'^ Archceologia, vol. Hi., p. 374. 


and pairs of animals in repousse work. The projections 
at each side of the rim, to which the crossbar at the 
top is attached, have pairs of human heads upon them. 
The mountings of the Aylesford bucket consist of three 
bronze bands, the lower two of which are plain and the 
uppermost one ornamented with pairs of animals and 
a peculiar kind of scrollwork. Each of the attachments 
for the handle at the top has upon it a single human 
head surmounted by a sort of crested helmet. 

The style of the art of the two buckets is the same, 
and corresponds in some respects with that of the 
Gaulish coins, ^ and in others with that of the sword- 
sheaths from La Tene"^' and the bronze situlce from 
Hallstatt, Watsch, and Certosa.-^ Dr. Arthur Evans 
has dealt so exhaustively with the details of these 
buckets and the origins of the art they exhibit in his 
paper on the Aylesford find in the Archceologia,^ that 
there is really little more to be said on the subject. It 
may, however, be worth while directing attention to 
the scrolls hanging down from the mouths of one of 
the pairs of beasts on the Marlborough bucket. Any- 
one unacquainted with the origin of these scrolls would 
probably mistake them for the animal's tongue pro- 
truding from its mouth, but on comparing the designs 
on the Marlborough bucket with those on the situlce 
just referred to, the scrolls will be seen to be simply 
degraded copies of the branch of a tree on which the 
animal is feeding. The art metalworkers who made 
the situlcB were, in fact, in the habit of using a simple 
convention for emphasising the difference between the 

^ Dictionnaire Archdologiqite de la Gaiile. 
■^ E. Voug-a, Les Helvhtes a La Tene. 

'^ Illustrated in the second edition of Dr. R. Munro's Boznia 
Herzegovina, p. 407. 
^ Vol. lii., p. 360. 


herbivorous and carnivorous animals by showing, in 
one case, the branch of a tree, and in the other the leg 
of its prey protruding from its mouth. The Celtic 
copyists were either ignorant of this convention or dis- 
regarded it, so that in their hands both the branches 
and leg's were soon converted into meaningless scrolls 
bearing hardly any resemblance to the original. 
Throughout the whole range of Celtic art there is 
displayed a tendency when dealing with plants and 
animals to transform first the details and afterwards the 
whole thing represented into curvilinear geometrical 

Besides the buckets just described, there are a few 
other examples of zoomorphic designs in Late-Celtic 
art, amongst which are the small bronze figures of 
animals found at Hounslow,^ Middlesex, now in the 
British Museum ; the bronze armlet, terminating in 
serpents' heads, from the Culbin Sands, '-^ Elginshire ; 
the knife-handle, terminating in a bull's head, from 
Birdlip,^^ Gloucestershire ; the iron fire-dogs,^ with up- 
rights terminating in horned beasts' heads, from 
Mount Bures, Essex, Hay Hill, Cambridgeshire, and 
Shefford, Bedfordshire; the horned bronze helmet from 
Torrs,^ Kirkcudbrightshire, now at Abbotsford ; the 
swine's head from Liechestown,*' Banffshire, now in the 
Banff Museum ; and the bull's head from Ham Hill,^ 
Somerset, in the Taunton Museum. 

The heads of the bull on the knife-handle from 

^ Proc. Soc. Ant. Loud., sen 2, vol. iii., p. 90. 

- Dr. J. Anderson's Scotland in Pagan Times: Iron Age, p. 156. 

■^ Trans. Bristol and Gloucestersh. Archceol. Soc, vol. v., p. 137. 

** Archceologia Camhrensis, ser. 6, vol. i., p. 41. 

^ Dr. J. Anderson's Scotland in Pagan Times: Iron Age, p. 113. 

^ Ihid.., p. 117. 

"' Proc. Somersetsli, Archceol. Soc, ser. 3, vol. viii., p. 2>3' 



Birdlip, and the beasts on the fire-dogs from Hay Hill 
and Shefford, have horns with round knobs on the end 
of each. Beasts with knobbed horns of this kind are 
represented on the Scandinavian gold bracteates^ of 
the Early Iron Age, generally associated with the 
swastika symbol. Similar horns are to be seen on the 
helmet of a small bronze figure found in Denmark;- 
on a figure depicted on the silver bowl from Gunde- 
strup,^ Jutland ; and on the handles of gold vessels 
from Ronninge,^ Boeslund,^ and Fyen,^ Denmark. 
The horns probably have some religious significance. 

Foliage so slightly conventionalised as to be easily 
recognised as such cannot be said to exist in Late-Celtic 
art, yet the foliageous origin of many of the designs at 
once betrays itself in the undulating curves with scrolls 
repeated at regular intervals on each side of what may 
be called the stem-line. We cannot select any better 
examples as illustrating this than the two beautiful 
bronze sword-sheaths from the crannog at Lisnacro- 
ghera,^ Co. Antrim. Here the portions of the designs 
which represent the principal stem consist of two 
lines running close together parallel to each other 
until they reach the point where a smaller stem branches 
off, when they diverge. The smaller stems, like the 
principal stem, consist of parallel lines running close 
together, and these, again, diverge to form what re- 

^ Prof. G. Stevens' Old Northern Runic Monuments. 

•^ J. J. A. Worsaae's Industrial Arts of Denrnarh^ p. 109. 

^ Sophus Miiller in Nordiske Fortidsminder, pt. 2, pi. 10. 

•* A. P. Madsen's Bronze Age, ii. , pi. 25. 

° Industrial Arts of Denmark, p. 105. 

^ P. B. Du Chaillu's Viking Age, vol. i., p. 97. 

" Jour. R. Hist, and A. A. oj Ireland, ser. 4, vol. vi., pp. 384-90. The 
decoration of a wooden tub found at the Glastonbury Marsh Yillag-e 
affords another very good instance of a Late-Celtic pattern derived 
from foliag-e. 











CO. ANTRi:,i ; 




presents the leaf. The ends of the leaves terminate in 
small spirals and their general shape resembles that of 
what are known as arabesques. We thus get the long 
sweeping S-shaped curves and the alternate contrac- 
tions and expansions of the space between the two 
boundary lines which are common to nearly all Late- 
Celtic ornament. Now, for some inscrutable reason, the 
natural forms of plant life never seem to have appealed 
to the Celtic mind in the way they did, for instance, to 
the ecclesiastical sculptors of the twelfth and thirteenth 
centuries. Consequently the designs which were in 
the first instance copied from foliage soon became 
transformed into a succession of beautiful flamboyant 
curves, pleasing to the eye unquestionably, but suggest- 
ing but little to the mind as to their meaning. In 
reference to this. Dr. Arthur Evans remarks : — ^ 

'*The tendency of all Late-Celtic art was to reduce all 
naturalistic motives borrowed by it from the classical world 
to geometrical schemes. . . . Yet the whole history of Late- 
Celtic art instructs us that this geometrical scheme, elaborate 
as it is, was originally based on ornaments of a naturalistic 

Once the foliageous origin of the flamboyant patterns 
was lost sight of or disregarded, it became easy to 
elaborate fresh designs by placing the forms derived 
from the leaves and stems of plants in all sorts of 
unnatural positions relatively to each other, as, for 
instance, on the pair of bronze spoon-like objects from 
Weston, near Bath, which are now in the Edinburgh 
Museum of Antiquities, and in a particular class of 
pierced ornaments, several of which are illustrated in 

^ Archceologia, vol. Iv. , p. 404. 



L. Lindenschmit's Die Alterthilmer unserer heidnischen 

Vorzeit. ^ 
A still further transformation resulted from the practice 

of drawing the various curves by means of a pair of 
compasses, and once this mechanical method 
had been adopted the temptation to intro- 
duce complete circles of different sizes into 
the designs would follow as a matter of 
course. This is very clearly seen on the 
ornamented bone spatulas from Slieve-na- 
Caillighe, Co. Meath, already referred to 
as having been found with a pair of iron 
compasses ; and also on backs of the bronze 
mirrors, of which a list is given on p. 115. 
It is most remarkable that the Late -Celtic 
artists should have succeeded in doing what 
has baffled everyone else before or since, 
namely, in producing ''sweet" curves by 
means of a combination of circular arcs. 

Lastly, when the patterns which had thus 
been evolved from natural foliage on a flat 
surface were transferred to the relief of the 
repousse metalwork, and raised bosses, 
volutes, and round plaques of enamel sub- 
stituted for the complete flat circles, an 
entirely new style of decoration was brought 
into existence. The most appropriate name 
that can be given to this particular kind 
of Celtic ornament \s flamboyant work. The 
French word flamhoyer means to blaze, and 

the Gothic window tracery of the fourteenth century, in 

which S-shaped curves predominate, is cdiWed flamboyant 

^ Vol. i., pt. X., pi. 6; vol. ii. , pt. viii. , pi. 5; vol. iii., pt. vii., pi. 6. 
Compare these with the ornament found at Silchester, illustrated in the 
ArchcBologia, vol. liv., p 470. 


Engraved Bone 
object from 
Co. Meath 






SCALE IT i.inea: 


on account of its resemblance to tongues of flame. 
The handle of the Late-Celtic tankard from Traws- 
fynydd,^ Merionethshire, now in the Liverpool Museum, 
if reproduced in stone on a larger scale, would certainly 
be mistaken for a piece of Gothic tracery, so that it may 
almost be looked upon as a blasphemous anticipation 
of Christian art by the Pagan Celt. 

The best examples of the Late-Celtic flamboyant 
work, for purposes of study, are the bronze shield 
from the river Thames, a circular disc of unknown 
use from Ireland, both in the British Museum ; the 
gold collar from Limavady, Co. Londonderry, in the 
Museum of the Royal Irish Academy, Dublin ; and 
the ^sica fibula, in the Newcastle Museum. There is 
also a disc in the Dublin Museum of similar desiofn to 
the one in the British Museum, which is worth com- 
paring with it. 

The whole design of the shield from the Thames is 
arranged with a due regard to symmetry. The small 
circular plaques of enamel, which are a leading feature 
in the scheme of decoration, are placed in definite 
positions, in groups of four and eight, around a central 
plaque within a raised boss. The plaques are con- 
nected by S-shaped curves in relief, which vary in 
width and in height above the background in different 
places. The highest part of the curve is emphasised 
by a sharp ridge which does not traverse the whole 
length of the curve midway between the margins, but 
at one place approaches near one edge, and a little 
further on approaches the other. An extremely com- 
plicated solid, bounded by curved surfaces, is thus 
formed, the appearance of which can only be realised 
by seeing the thing itself or a model of it. 

^ Arcliceologia Cmnhrensis, ser. 5, vol. xiii., p. 212. 



The circular bronze discs and the gold collar from 
Ireland exhibit a class of flamboyant work which is 
somewhat different from that on the shield from the 

Hbula of Bronze Gilt from ^Esica 
Now in the Newcastle Miiseiini. Scale j linear 

Thames, and is still further removed from the original 
foliage motive designs. Here conchoids take the place 
of the circular enamelled plaques arranged in sym- 
metrical positions, and the curves connecting them 





with each other have the trumpet-shaped terminal 
expansions, which are so characteristic of the Late-Celtic 
style, very highly developed. A 
further modification that disguises 
the foliageous origin of the design 
is the substitution of two C-shaped 
curves meeting at an angle for 
the more gracefully flowing S- 
shaped curves. Examples of a 
running pattern composed of C- 
shaped curves meeting at an angle 
in the way described, occur on a 
bronze collar from Lochar Moss,^ 
Dumfriesshire, now in the British 
Museum (see p. 112). A running 
pattern composed of C and S- 
shaped curves alternately meeting 
at an angle occurs on the enamelled 
mounting of a bronze bowl from 
Barlaston,'-^ Staffordshire. We have 
pointed out the changes due to 
copying in relief a design en- 
graved on a flat surface ; but 
curiously enough when the decora- 
tion of the repousse metalwork 
was again transferred to a flat 
surface, as in the enamelled fittings 
of the bronze bowls and in the 
spiral ornamentation of the illum- 
inated MSS. of the Christian 
period, it did not return to what 
it was before, but became still 





^ ArchcBologia , vol. xxxiii. 
" Ibid., vol. Ivi., p. 44. 

P- 347- 

Flamboyant Ornament on 

Collar from Broig-hter, 

Limavady, Co. Londonderry 



more unlike its foliageous prototype. It will be noticed 
that the ends of the trumpet-shaped expansions on the 
bronze discs and gold collar being in the highest relief 
catch the light. In the MSS. and enamels this effect 
is imitated by 
small almond- 
shaped spots of a 
different colour 
from the rest. 
The beautiful 
repousse orna- 
ment on the 
bronze mirror 

from Balmaclel- Spiral Ornament in Illuminated MS. 

Ian, ^ Kirkcud- copied from repousse metalwork 

brightshire, now in the Edinburgh Museum of Anti- 
quities, supplies us with another instance of little raised 
bosses which were afterwards reproduced on the flat by 
means of colour in the Christian MSS. 

Many of the curvilinear Late-Celtic patterns which 
are used to fill a circular space are based upon the 
triskele and the swastika. A good example of a curved 
swastika pattern occurs on each of the three enamelled 
handles of a bronze bowl found at Barlaston,'^ Stafford- 
shire. Triskele designs are much more common, 
especially on the round disc fibulae, specimens of which 
have been found at Silchester," Hampshire ; Brough,"* 
Westmoreland ; and in the Victoria Cave,'* near Settle, 
Yorkshire. There are other instances on the bronze 

^ Dr. J. Anderson's Scotland hi Pagan Times: Iron Age, p. 127. 

^ Archceologia, vol Ivi. , p. 44. 

^ Now at Strathfieldsaye House. 

^ Proc. Soc. Ant. Lond., ser. i, vol. iv. , p. 129. 

^ Historic Soc. of Lane, and Cheshire, Trans, for 1866, p. 199. 





tankards from Elveden,^ Essex, and Trawsfynydd,'- 
Merionethshire; on the bronze shield from the Thames,^ 
now in the British Museum; on a bronze disc-and-hook 
ornament in the Dublin Museum;^ on a bronze plate 
in the Welshpool Museum;^ on some bronze harness- 
mountings (?) from South Shields;^ on bronze wheel- 
shaped pendants from Seamill Fort,' Ayrshire, from 
Berkshire, now in the British Museum, from Kings- 
holm,^ near Gloucester, and from Treceiri, Carnarvon- 
shire. These designs may have had a symbolical 
origin, as the triskele was a well-known sun symbol 
in the Bronze Age. The triskele arrangement of three 
spirals round a central spiral survived in the decoration 
of the illuminated MSS. of the Christian period. 

In the repousse metalwork of the Late-Celtic period 
certain portions of the design are thrown into relief in 
order to enable them to be distinguished from the rest 
which is not in relief. Much the same artistic effect 
can be obtained when the design is engraved on a 
flat surface by means of shading, and in the case of 
enamelled plaques, by employing different colours. 
In fact, by the use of relief, shading, or colour, the 
decorative effect of a pattern is doubled, because there 
are two things for the mind to comprehend, namely, 
the shape of the pattern itself and the shape of the 
background. Anyone who endeavours to realise both 
shapes simultaneously will find it an impossibility. 

Several different kinds of shading are used in Late- 

^ ArchcEologia , vol. lii., p. 359. 

■^ Archceologia Canibrensis, ser. 5, vol. xiii. , p. 212. 

■^ Kemble's Horce Ferales, pi. 15. 

■* The Reliquary for 190 1, p. 56. ^ Unpublished. 

^ Jour. Brit. Archceol. Assoc, vol. xxxix. , p. 90. 

"^ R. Munro's Prehistoric Scotland, p. 378. 

^ Doug-las' Nenia Britannica p. 134. 


Celtic art, chiefly in ornament engraved on metal, wood, 
bone, and pottery, as will be seen by the following 

List showing different kinds of shading used in Late-Celtic 
Art^ and the objects on which they occur. 

(i) Shading of parallel lines. 

On spoon-like bronze objects from Crosby Ravensworth, 
Westmoreland, and Ireland. 

On bronze mirror from Stamford Hill, near Plymouth. 
On engraved pottery from Glastonbury Marsh Village. 
On bronze sword-sheath from Embleton. 

(2) Cross-hatching placed diagonally. 

On engraved piece of wood and engraved pottery from 
the Glastonbury Marsh Village. 


(3) Cross-hatching placed diag-onally, with dots in each of 
the square meshes. 

On engraved wooden tub from the Glastonbury Marsh 

(4) Cross-hatching of double lines placed diagonally. 

On engraved piece of wood from the Glastonbury Marsh 

(5) Chequerwork grass-matting shading. 




On bronze sword-sheath from crannog- at LIsnacroghera. 

On bronze mirrors from Trelan Bahow, Cornwall ; Birdlip, 
Gloucestershire ; from unknown locality, now in the Liver- 
pool Museum ; and from Stamford Hill, near Plymouth. 

(6) Engine-turned shading. 

On gold collar from Limavady. 
(7) Dotted shading. 

On bronze spoon-like objects in the Dublin Museum. 

On bronze harness-ring from Polden Hill, Somersetshire. 

On silver armlet from Stony Middleton, Bucks. 

Besides the Late-Celtic objects just described, which 
exhibit curvilinear surface decoration derived from 
foliage, there are others with very peculiar forms ''in 
the round." Amongst these are the harness-rings with 
projecting knobs from Polden Hill, Somersetshire ; 


Stanwick, Yorkshire, and elsewhere ; the beaded 
torques from Lochar Moss, Dumfriesshire ; Hynford, 
Lanark ; and the beaded bracelets from Arras and 
Cowlam, Yorkshire. 

The projections on the harness-rings generally occur 
at three points round the circumference, and their 
shapes will be better understood from the illustrations 
than from any written description. It is not easy to 
say what the meaning or origin of these projections can 
be, as they bear no obvious resemblance to any natural 
or artificial object. 

The beaded torques mentioned are composed of 
separate metal beads (usually of two different shapes) 
strung on a square iron rod, so that they cannot rotate 
or rattle about. The bracelets are, however, cast in 
one piece, and made in imitation of a string of beads. 
This style of bracelet is of foreign origin, as specimens 
have been found in France^ and Germany,'^' many of 
which are elaborately ornamented with spiral-work in 
high relief. 

Rectilinear patterns are of comparatively rare occur- 
rence in Late-Celtic art, as the designers of the period 
appear to have had a rooted objection to using straight 
lines if they could possibly be 
avoided. There are, however, a 
few exceptions. The small circular 
enamelled plaques with which the 
bronze shield from the Thames, 
now in the British Museum, is 
decorated, have a swastika pattern 
on each. The swastika was pro- 
bably a foreign importation, as it Swastika desig-n on Shield 

from the Thames 
^ Dictionnaire Archdologique de la Gaiile. 
'^ Lindenschmit's Alterthilmer. 


is used in the decoration of the Gaulish bronze helmet 
from Gorge-Meillet^ (Marne), and of the iron lance- 
head from La Tene,- Switzerland. 

The step-pattern in Late-Celtic art may have had a 
textile origin, i.e, have been copied from a woven belt 
or other fabric. Instances of it occur on a piece of 
engraved wood from the Glastonbury^ Marsh Village ; 
on the bronze mountings of a shield from Grimthorpe,^ 
Yorkshire, now in the British Museum ; on the bronze 
ferrule of a spear-shaft from the Crannog at Lisnacro- 
ghera,'^ Co. Antrim ; and on a sculptured monolith 
at Turoe, Co. Galway. The step- pattern survived 
after the Pagan period in the Christian enamels, as in 
the bowl from Moklebust,^ Norway, and the fragment 
at St. Columba's College,^ Dublin. The key-pattern, 
or Greek fret, is unknown in Late-Celtic art. 

The chequerwork pattern may also have had a textile 
origin. There is an example of it on the bronze 
sword-sheath from Embleton,*^ Cumberland, now in 
the British Museum. 

The chevron and lozenge patterns are possibly sur- 
vivals from the preceding Bronze Age. We have 
instances of the chevron pattern on the bronze mirror 
from Trelan Bahow,'' Cornwall, and on a potsherd from 
the Glastonbury 1" Marsh Village ; and of the lozenge 
on the stave of a bucket^^ from the same site. 

^ A. Bertraiid's ArchMogie Celtigue et Gaiiloisc, p. 367. 

2 E. Voug-a's Les Helvetes a La Tene, pi. 5. 

•^ The Antiqiiary ioY 1895, P- ^^O- 

■* LI. Jewitt's Grave Monnds and their Contents, p. 246. 

■^ Jour. R. Hist, and Archceol. Assoc, of Ireland, sen 4, vol. vi. , p. 394. 

*^ Mdm. de la Soc. Ant. du Nord, 1890, p. 35. 

" J. B. Waring"'s Manchester Fine Art Treasures Exhibition. 

^ Kemble's Horce Ferales, pi. 18, fig". 3. 

'^ Archceol. Jour. ^ vol. xxx. , p. 267. 
•'*' Proc. Somersetsh. Archceol. Soc, vol. xl. 
•'^ The Afitiquary ioY 1895, p. 110. 



(a.d. 450 to 1066) 


IT must always be borne in mind that the conversion 
of the inhabitants of Britain from Paganism to 
Christianity was a very gradual process, extending 
over a period of two hundred years at least. It seems 
probable that during the last hundred years or so of 
the Roman occupation of Britain the Christian faith 
may have been accepted by a limited number of the 
native population ; but almost as soon as the new 
religion began to take root in England it was entirely 
swept away by the Saxon conquest, and the few con- 
verts who were not exterminated by the ruthless Pagan 
invaders fled for refuge to Wales and Cornwall. The 
archaeological evidence of the existence of Romano- 
British Christianity is extremely scanty. Out of the 
hundreds and hundreds of inscribed and sculptured 
monuments belonging to the period of the Roman 
occupation of Britain there is not one which bears 
a Christian symbol or shows a trace of Christian art. 
There are only two instances of the occurrence of a 
Christian symbol on a Romano-British structure, 
namely, (i) at Chedworth,' where the Chi-Rho Mono- 

^ journ. Brit. ArchceoL Assoc, vol. xxiii., p. 228. 



gram is carved twice upon a stone in the foundation 
of the steps leading into the corridor of a Roman villa 
there; and (2) at Frampton/ Dorsetshire, where the 
same Monogram forms part of the decoration of a 
mosaic pavement in one of the rooms of a Roman 
villa. As Romano-British Christianity produced no 
effect on the art of this country, we are not further 
concerned with it. 

Whilst England remained under the dominion of 
Saxon Pagandom for a century and a half in some 
parts, and for nearly two centuries in others, Christianity 
spread rapidly from Gaul to Cornwall, Wales, and the 
south-west of Scotland, and thence to Ireland. After 
the Saxons were converted by St. Augustine, in 
A.D. 597, there was a return wave of Celtic Christianity 
from Ireland to lona, and from lona to Lindisfarne, 
in Northumbria, which was founded a.d. 635. The 
localities where Christianity was first planted in Britain 
are indicated archseologically by the geographical 
distribution of monuments bearing the Chi-Rho Mono- 
gram, which is as follows : — 

Cornwall. St. Just. 

St. Helen's Chapel. 


Carnarvonshire. Penmachno. 
Wigtownshire. Kirkmadrine. 


As the Chi-Rho Monogram does not occur on the 
early inscribed stones of Ireland, but in place of it the 
cross with equal arms expanded at the ends, enclosed 
in a circle, which is derived from the Monogram,'^ it 

^ S. Lysons' Reliquice Brittanico RomaiKE, No. 3, pi. 5. 
^ See J. R. Allen's Christian Symbolism, p. 94. The Chi-Rho Mono- 
gram occurs on inscribed monuments in Gaul between a.d. 377 and 493. 


naturally follows that Irish Christianity is later than 
that of Cornwall, Wales, and the south-west of 

Setting aside the vague and unsatisfactory statements 
of the mythical period (such as the one about the 
presence of three British at the Council of Aries in 
A.D. 314), we find that the real history of the Christian- 
ising of this country begins with the opening years of 
the fifth century, and that it followed directly from the 
foundation of the school of learning and centre of 
missionary enterprise by St. Martin at Tours, in France. 
In A.D. 397 St. Martin died, and not long after, in 
A.D. 412, his disciple, St. Ninian, built a stone church 
dedicated to his master at Whithorn, Wigtownshire. 
In A.D. 429 Germanus, Bishop of Auxerre, and Lupus, 
Bishop of Troyes, visited Britain in order to suppress 
the Pelagian heresy. About the same time the con- 
version of Ireland is believed to have been commenced 
by either St. Patrick or by St. Palladius [circa a.d. 432). 
The sixth century witnessed the foundation of the great 
school of ecclesiastical learning at Llantwit Major, 
Glamorganshire, where St. David, St. Samson, and 
Gildas the historian were educated ; but an event of 
even greater importance was the landing of St. Columba 
at lona in a.d. 563, and the subsequent conversion of 
the northern Picts. The sixth century ends with the 
conversion of Kent by St. Augustine in a.d. 597. It 
was eighty-four years more before the South Saxons 
accepted Christianity and the conversion of England 
became complete. In the meantime the difi'erences 
between the Saxon and Celtic Churches had been 
settled in favour of the former at the Synod of Whitby 
in A.D. 664. 

Reviewing the historical facts just mentioned, it 



appears that for about 200 years (from a.d. 450 to 650) 
there was a separate Celtic Church in Britain, which 
may appropriately be called the pre-Augustinian 
Church. The question now naturally suggests itself, 
to what extent did the introduction of Christianity 
influence the native art of Britain during the 200 years 
which followed the departure of the Romans from its 
shores? The answer supplied by archaeology is that 
before about a.d. 650 there was no distinctively Chris- 
tian art existing in this country. 

The monuments belonging to the pre-Augustinian 
Church consist of rude pillar-stones with incised crosses 
of early form, or with Latin inscriptions in debased 
Roman capitals, sometimes with Celtic inscription in 
Ogams in addition. The monuments of this class do 
not, as a rule, show any trace of ornament or sculpture 
beyond the crosses and inscriptions. The only recorded 
exceptions are — 

An Ogam-inscribed stone from Pentre Poeth,i Brecknock- 
shire, now in the British Museum, having on one face a 
bishop with his crozier, St. Michael and the Dragon, and 
very rude zigzag ornament. 

An Ogam-inscribed stone from Glenfahan,^ Co. Kerry, 
now in the Dublin Museum, with rude spiral ornament, a 
figure of a man, a looped pattern, and several crosses. 

An Ogam-inscribed stone at Killeen Cormac,^ Co. Kildare, 
lying prostrate near the entrance gate, with a bust of Christ 
carrying the cross over the right shoulder. 

St. Gobnet's Stone at Ballyvourney,* Co. Cork, with a 
cross enclosed in a circle, surmounted by the figure of a 
bishop holding his crozier. 

^ Archceologia Cambrensis, sen 6, vol. i. , p. 240. 

'■^ Trans. Royal Irish Academy, vol. xxxi., p. 318. 

•^ Journ. R. Hist, and A. A. of Ireland, sen 4, vol. ii., p. 546. 

^ Archceol. Jour., vol. xii., p. 86. 

1 66 


A stone, with a minuscule inscription, at Reask,i Co. Kerry, 
having on the same face a cross in a circle, with incised spiral 
ornament at each side of the shaft. 

The stones, with incised symbols of unknown mean- 
ing, which are so common in the north-east of Scotland, 

Enamelled Handles of Bronze Bowl found at Barlaston, Staffordshire 

Now in the possession of Miss Amy Wedgwood. Scale } linear 'j, 

possibly belong to the same early period. The orna- 
ment on some of them has a very marked Late-Celtic 

There are no Celtic MSS. with illuminations or orna- 
ment of any kind to which a date earlier than a.d. 650 

■^ Archceologia Camhrensis, sen 5, vol. ix. , p. 147. 


can be assigned, but there are a certain number of 
metal objects which illustrate the overlap of the Pagan 
and Christian styles of Celtic art. Amongst the most 
important of these are the bronze bowls with enamelled 
mountings and zoomorphic handles which have been 

Enamelled Handle of Bronze Bowl from Chesterton-on-Fobsway, 

Now in the Warwick Museum. Scale } linear 

described at some length by the author in the Archceo- 
login (vol. Ivi., p. 43). The chief peculiarities of the 
bowls is the hollow moulding just below the rim and 
the three or four handles with rings for suspension. 
The upper part of each handle is like a hook, termi- 
nating in a beast's head, which rests on the rim of the 
bowl and projects inwards over it. The lower part of 



each handle is circular, or in the shape of the body of 
a bird, and is fixed to the convex sides of the bowl. 
The circular form is most common in the examples 
found in England, and the disc is either ornamented 
with cliampleve enamel ^ or with piercings, giving a 
cruciform appearance. -' 

Enamelled Handle of Bronze Bowl from Chesterton-on-Fossway, 

Now in the Warwick Museum. Scale | linear 

The earliest of the series from Barlaston, Stafford- 
shire, now in the possession of Miss Wedgwood, has 
three handles all alike, ornamented with discs of enamel, 

^ As in the specimens from Barlaston, Staffordshire ; Chesterton-on- 
the-Fossway, Warwickshire ; Barring-ton, Cambridgeshire; Crosthwaite, 
Cumberland ; Middleton Moor, Derbyshire ; Oxford ; and Greenwich. 

- As in the specimens from Wilton, Wilts ; and Faversham, Kent. 


the designs on which are distinctly Late-Celtic in 
style, and consist of small circles connected by C- and 
S-shaped curves. In the case of the enamelled handles 
of the other specimens, closely coiled spirals of the 
Bronze Age type take the place of the circles, and by 
this trifling alteration the character of the design is so 
completely changed as to be almost identical with the 
spiral decoration of the Book of Durrow and other 
Irish MSS. of the same period. We see here exactly 

Spiral Ornament from the Book of Durrow 

when and how the flamboyant ornament of Pagan 
Celtic art became transformed into the spiralwork of 
the Christian illuminated MSS. which was afterwards 
applied to the decoration of the sculptured crosses and 
ecclesiastical metalwork. The circumstances under 
which the bowls have been found show that they be- 
long to the Pagan Saxon period between a.d. 450 
and 600. 

In the museum of the Society of Antiquaries at 
Burlington House there is the cast of an object from 


the collection of Mr. Albert Way, the well-known 
antiquary, which exhibits a curious mixture of styles. 
Where the original is, or where it came from, is unfor- 
tunately not known, but it has every appearance of 
having been of metal. In the middle of the object is a 
square panel of triangular pierced work, exactly like 
that on the cover of the Stowe Missal^ (made a.d. 1023 
to 1052) ; whilst at each of the rounded ends are curved 
designs with trumpet-shaped expansions of pronounced 
Late-Celtic type. 

Plaitwork, which is, of course, one of the leading 
motives of Celtic art of the Christian period, occurs 
occasionally in association with Pagan flamboyant 
ornament, as on a brooch from the Ardakillen"^ crannog, 
near Stokestown, Co. Roscommon (now in the Dublin 
Museum), and on a gold armlet from Rhayader,^ 
Radnorshire (now in the British Museum). 

Amongst objects belonging to the early Christian 
Celtic period before a.d. 600, may probably be classed 
the leaf-shaped silver plates engraved with symbols 
from Norrie's Law,"^ Forfarshire, and the terminal link 
of a silver chain, also engraved with symbols, from 
Crawfordjohn,^ Lanarkshire (all in the Edinburgh 
Museum of Antiquities). The hammer-headed pins 
also, a list of which has already been given (p. 108), 
seem, from the enamelled designs upon them, to belong 
to the transitional period between Celtic Paganism and 

Although, as we have just seen, the introduction of 

^ Miss M. Stokes' Early Christian Art in Ireland, p. 92. The Stowe 
Missal is in the Museum of the R.I. A. at Dublin. 
- Sir W. Wilde's Catal. Mus. R.I. A., p. 569. 

'•' ArchcEologia Camhretisis, ser. 5, vol. xvi., p. 261. 

■* Dr. J. Anderson's Scotland in Early C/jristian Times, z\\(\ ser., p. 38. 

■' Ibid.., p. 44. 






(a.U. 1023 TO 1052) 


Christianity into Britain did not immediately affect the 
native Pagan art to any appreciable extent, yet as soon 
as the Saxons were converted and communication with 
the Continent became easier and therefore more frequent, 
an entirely new style of decoration came into existence 
with extraordinary rapidity. The flamboyant designs 
of the Late-Celtic period were modified by combining 
them with the closely coiled spiral of the Bronze Age, 
and several new motives, such as interlaced-work, key- 
patterns, zoomorphs, and foliage, were introduced from 
foreign sources. At the same time a complete revolution 
took place in the class of objects to the decoration of 
which the skill of the artificer was applied. The priest 
took the place of the warrior as the patron of the fine 
arts, and monopolised all the available time of the 
metalworker and enameller in making beautiful vessels 
for the service of the church. Then, too, with Chris- 
tianity came the art of writing and illuminating 
ecclesiastical MSS., which was unknown to the Pagan 
Celt. The influence of the draughtsman upon other 
arts was now possible for the first time, and the intro- 
duction of MSS. soon worked far-reaching changes. 
Fresh motives could be more easily transferred from 
one art centre to another, and decorative designs could 
be combined and elaborated in a way that was impossible 
when working in such intractable materials as metal or 
stone instead of drawing on parchment with a facile 
pen. The new Celtic style of the Christian period soon 
took a definite shape, and after the patterns had been 
uUy developed in the illuminated MSS. they were 
afterwards applied to decorative work in stone and 



The materials available for the study of Celtic Art of 
the Christian period may be divided into four classes, 
namely : — 

(i) Illuminated MSS. 

(2) Sculptured Stones. 

(3) Metalwork. 

(4) Leathervvork, Woodwork, and Bonework. 

The most important collections of Irish and Hiberno- 
Saxon MSS. in this country are in the libraries of 
Trinity College, Dublin ; of the Royal Irish Academy, 
Dublin ; and the British Museum, London. There are 
other smaller collections, or in some cases single 
volumes only, in the University and College libraries 
of Oxford and Cambridge ; in the Cathedral libraries 
at Durham, Lichfield, and Hereford ; and in the 
Archiepiscopal library at Lambeth. The chief libra- 
ries on the Continent which are fortunate enough to 
possess specimens of Irish calligraphy and illumination 
(either acquired by purchase or still the property of 
monasteries originally founded by Irish missionaries) 
are at Stockholm, St. Petersburg, Paris, St. Gall and 
Basle in Switzerland, and at Nuremberg, Fulda, and 
Treves in Germany. The Irish MSS. from the monas- 
tery founded by St. Columbanus in a.d. 613 at Bobio, 
in Piedmont, are distributed over the libraries at Milan, 
Turin, and Naples. For descriptions and illustrations 
of these MSS. the reader may be referred to Prof. J. O. 
Westwood's Palceographia Pictoria Sacra and Minia- 
tures of the Anglo-Saxon and IrisJi MSS.; C. Purton 
Cooper's Report on Rymer's Foedera^ Appendix Ay 


Sir H. James' Facsimiles of the National MSS, of Ire- 
land; Publications of the Palceographical Society ; Miss 
Margaret Stokes' Early Christian Art in Ireland ; Dr. 
J. Stuart's Book of Deer^ published by the Spalding 
Club of Aberdeen ; J. A. Bruun's Illuminated Manu- 
scripts of the Middle Ages ; and Dr. W. Reeve's paper 
on ''Early Irish Caligraphy " in the Ulster Journal of 
Archceology, vol. viii., p. 210. 

The following is a list of Irish MSS. selected on 
account of the beauty of their illuminated pages : — 


Book of Lindisfarne 
Book of Kells . 
Book of Durrow . 
Book of Armagh 
Book of St. Chad 
Book of MacRegol 
Book of MacDurnan 
Book of Deer 
Codex No. 51 
Golden Gospels . 
Gospels of St. Arnoul, 

British Museum (Nero D. iv. ). 

Trinity College, Dublin. 


Royal Irish Academy, Dublin. 


Bodleian, Oxford. 


Public Library, Cambridge. 

St. Gall, Switzerland. 

Royal Library, Stockholm. 

Imperial Library, St. Peters- 

Metz Nuremberg-. 



Vespasian A. i. . . . British Museum. 

Vitellius F. xi. . . . Ibid. 

Psalter of St. John's College Cambridge. 

Psalter of Ricemarchus . Trinity College, Dublin. 

Some of the above MSS. can be dated by means of 
entries giving the name of the scribe or other person, 
who can be identified by contemporary or nearly con- 
temporary historical record. The oldest MS. with 


illuminations in the Hiberno-Saxon style which can 
be thus dated is the Lindisfarne Book. It contains 
two entries written in an English hand of the tenth 
century, w^iich show that the volume was written by 
Eadfrith, Bishop of Lindisfarne ; that ^thilwold, 
Bishop of Lindisfarne, made the cover for it ; that 
Billfrith, the anchorite, wrought the metalwork for it ; 
and that Aldred, the priest, over-glossed it in English 
for the love of God and St. Cuthbert. Eadfrith held 
the see of Lindisfarne from a.d. 698 to 721, and was 
then succeeded by ^Ethilwold, who held the bishopric 
of the island until his death in a.d. 740. The Book of 
Lindisfarne, therefore, must have been written either 
during the last two years of the seventh century or the 
first twenty-one years of the eighth century. This 
may be looked upon as the starting-point of all 
Hiberno-Saxon art, and its origin may be fairly traced 
to Lindisfarne, where the Scotic and Anglo-Saxon 
schools were able to mingle, each reinvigorating the 
other to their mutual advantage. 

The Book of Kells makes its first appearance in 
history in a.d. 1006, during which year it is recorded 
in the Annals of the Four Masters that the Great 
Gospels of Columkille was stolen. Although the name 
of the scribe who wrote and illuminated this book is 
unknown, it is probable, from the style of the decora- 
tion and lettering, that it belongs to about the same 
period as the Lindisfarne Book, but somewhat later, 
as the Book of Kells contains foliage amongst the 
ornament, and is altogether more elaborate. 

The Book of Durrow was written by a scribe named 
Columba, who can hardly have been the celebrated 
Saint of that name, as his time is far too early for it. 
Since the spiral patterns in the Book of Durrow 


approximate more nearly to the flamboyant designs of 
the Pagan-Celtic metalwork than those in any other 
MS., it cannot be dated later than the eighth century. 

The Book of St. Chad should more properly be 
called the Book of St. Teilo, as it contains an entry 
stating that the volume was purchased by Gelhi, son of 
Arihtuid, from Cingal for his best horse, and dedicated 
to God and St. Teilo. Before it was at Lichfield it lay 
on the altar of Teilo, at Llandaff. This MS. has also 
a good claim to be of the eighth century. 

The Book of Armagh and the Golden Gospels of 
Stockholm are of the ninth century. The former was 
written by Ferdomnach, ''a sage and choice scribe of 
Armagh," who died in a.d. 844. The Stockholm 
Gospels contains a deed of gift, which shows that the 
volume was bought by the Earl Alfred and Wetburg 
his wife from a Viking, and presented by them to the 
Cathedral of Canterbury. The deed is signed by 
JElired, Wetburg, and their daughter Alhtryth, who 
have all been identified by the will of Alfred, which is 
attested by ^Idered, Archbishop of Canterbury, from 
A.D. 871-9. The Gospels of MacRegol also belongs 
to the ninth century, if the identification of the scribe 
who wrote it with " MacRiagoil nepos Magleni, Scriba 
et Episcopus Abbas Biror" can be relied upon. His 
death is recorded in the Irish Annals under the year 
A.D. 820. 

The Gospels of MacDurnan is of the tenth century. 
It has an inscription on one of the blank pages of the 
MS. showing that the book was either written for, 
or was in the possession of, Maelbrigid MacDurnan, 
and that it was given by King Athelstan to the city 
of Canterbury. Maelbrigid MacDurnan was Abbot 
of Derry in the ninth century, and was afterwards 


promoted to the see of Armagh in a.d. 927. He died 
in A.D. 927. Athelstan reigned from a.d. 925 to 941. 

The Psalter of Ricemarchus is of the eleventh 
century. It contains a Latin poem, from which we 
gather that the book was written by Ricemarch 
Sulgenson, with the assistance of Ithael, ''whose 
name makes learning golden," and that the initial 
letters were illuminated by John. Ricemarch, or 
Rhyddmarch, succeeded his father Sulgen in the see 
of St. Davids in a.d. 1089, and died in a.d. 1096. 

The examples given afford a very good series 
arranged in chronological order, showing the modifi- 
cations which the style underwent in the course of the 
four centuries between a.d. 650 and 1050. We are 
somewhat sceptical as to there having been any fine 
illuminated Hiberno-Saxon MSS. before a.d. 700; but 
assuming that there may have been some which are no 
longer in existence, the best period is from a.d. 650 to 
850 ; then from a.d. 850 to 950 there is a middle period 
of rather inferior excellence; and, lastly, from a.d. 
950 to 1050 a distinct period of decline which went on 
with increasing decadence for a century or two after 
the Norman Conquest. 

The number of illuminated pages in the different 
MSS. varies considerably, sometimes because the 
volumes are imperfect, but also because they were 
less lavishly illustrated in the first instance. The illu- 
minated pages in the copies of the Gospels are of the 
following kinds : — 

(i) Initial pages. 

(2) Ornamental or Cross pages. 

(3) Symbols of the Evangelists. 

(4) Portraits of the Evangelists. 

(5) Scenes from the Life of Christ. 

(6) Tables of Eusebian Canons. 


As an instance of a very completely illustrated MS. 
of the Gospels we may take the Lindisfarne Book, which 
contains twenty-three full pages of illumination as 
specified below : — 

Four portraits of the Evangelists with their Symbols, one 

for each Gospel. 
Five ornamental pages, one before St. Jerome's Epistle 

and one before each Gospel. 
Six Initial pages ^ namely — 

" Novum opus," commencing St. Jerome's Epistle. 

''Liber generationis," commencing St. Matthew's 

" XPI autem generatio," commencing the Genealogy 
of Christ in St. Matthew's Gospel. 

" Initium Evangelii," commencing St. Mark's Gospel. 

" Quoniam quidem," commencing St. Luke's Gospel. 

" In principio erat," commencing St. John's Gospel. 
Eight pages of tables of Eusebian Canons. 

The Book of Durrow has sixteen illuminated pages, 
namely, four of the Symbols of the Evangelists ; six 
ornamental pages, one at the frontispiece, one before 
the Preface of St. Jerome, and one before each Gospel ; 
and the usual six initial pages. 

The Book of Kells is more profusely illustrated than 
any other Irish MS. in existence. Besides innumerable 
large and small initials, it contains three portraits of 
the Evangelists, three combined symbols of the Four 
Evangelists, three scenes from the Life of Christ — 
namely, the Virgin and Child, Christ seized by the 
Jews, and the Temptation of Christ, and eight pages 
of Eusebian Canons. 

The St. Gall Gospels (Codex No. 51) has twelve full 
pages of illumination, namely, four portraits of the 
Evangelists, five initial pages, one ornamental cross- 



page, and two scenes from the Life of Christ — the 
Crucifixion and Christ in Glory. 

As an instance of the method of illustrating the 
Irish MSS. of the Psalter we may take the one in the 
library of St. John's College, Cambridge. This has 
six illuminated pages, namely- — ■ 

(i) " Beatus vir," commencing the ist Psalm. 

(2) ''Quid gloriaris" ,, ,, 51st 

(3) " Doe exaudi " ,, ,, loist ,, 

(4) Miniature of the Crucifixion. 

(5) ,, David and Goliath. 

(6) ,, David and the Lion. 

The Vit. F. xi. Psalter in the British Museum has 
two initial pages and two miniatures, namely, David 
and Goliath, and David playing the harp. 

The Vesp. A. i. Psalter in the British Museum has 
only one miniature, namely, David playing the harp ; 
but it has a great number of extremely beautiful initial 
letters ornamented with spiralwork of the best quality. 
Figure subjects (one of David and the Lion) are 
introduced in the initials of the 26th, 52nd, 68th, 97th, 
and 109th Psalms. 

The details of the ornamental patterns in the MSS. 
will be dealt with when we come to consider the leading 
characteristics of the style ; all that we need do now, 
therefore, is to point out the manner in which the 
patterns are distributed. The treatment of the minia- 
tures of the Evangelists and of the scenes from the 
Life of Christ and the Life of David is very simple ; 
the picture is enclosed within a rectangular frame 
divided into panels, each filled in with a separate piece 
of ornament complete in itself. Sometimes, as in the 
case of the miniatures of Christ seized by the Jews in 
the Book of Kells, and David playing the harp in the 


Vesp. A. i. Psalter, the figures are placed beneath an 
arch supported by columns at each side. The archi- 
tectural origin of the design is entirely concealed by 
converting the columns and the arch into pieces of flat 
ornament arranged in panels. The pages of Eusebian 
Canons are also treated architecturally, the tables being 
placed under arcading so disguised by the incrustations 
of ornament as to be almost unrecognisable. The 
initial pages of the Gospels are only partially sur- 
rounded by a rectangular frame, so as to allow the 
tops of the large capital letters to project beyond the 
frame into the margin. The incomplete portion of the 
frame on the right side of the page is coverted into a 
zoomorph in a characteristically Celtic manner by 
adding the head of a monster at the top and a fish-like 
tail at the bottom. The frame and the larger initials 
within it are covered with panels of ornament. The 
pages of ornament are generally arranged in rect- 
angular panels, so as to give the appearance of a 
cross ; or sometimes, as in the Book of Durrow, there 
is a small equal-armed cross within a circle in the 
middle of the page, the remainder of which is entirely 
filled up with ornament. In many cases where the 
miniatures, etc., are surrounded by a rectangular 
frame the outer margins are extended and formed into 
ornamental knots at each of the four corners. 

After the Celtic style of decorative art of the Chris- 
tian period had been fully developed in the Irish 
and Hiberno-Saxon illuminated MSS. of the eighth 
century, it was afterwards applied to sculptured stone- 
work in the ninth, tenth, and eleventh centuries. There 
are so few details of pre-Norman Celtic buildings^ 

^ The sculptured architectural details of the Round Towers and early 
churches in Ireland and Scotland consist chiefly of crosses or crucifixes 
over the doorways and terminal heads. 


which afford examples of ornamental sculpture that 
they are hardly worth considering, so that we need 
only take cognisance of the sepulchral and other 
monuments. These are of the following different 
kinds : — 

(i) Recumbent cross-slabs, 

(2) Recumbent hog-backed and coped stones. 

(3) Erect cross-slabs. 

(4) Erect wheel-crosses. 

(5) Erect free-standing crosses. 

(6) Erect pillar crosses, with shafts of round or square 


The recumbent cross-slabs are confined almost ex- 
clusively to Ireland, although there are one or two in 
Cornwall, Wales, and Scotland. Much the largest 
collection is at Clonmacnois, King's Co., where there 
are not far short of 200 sepulchral cross-slabs with 
inscriptions in Irish minuscule letters, giving the name 
of the deceased and requesting a prayer for his or her 
soul. A considerable number of the names on the 
slabs have been identified on sufficiently satisfactory 
evidence, thus giving reliable dates for a series 
arranged in chronological order. Clonmacnois was 
founded by St. Ciaran in a.d. 554, but the greater part 
of the dated cross-slabs belong to the ninth, tenth, and 
eleventh centuries. The earliest of these inscribed 
cross-slabs which exhibits any decorative features is 
that of Tuathgal,^ who has been identified with the 
seventh abbot of Clonmacnois. The death of abbot 
Tuathgal took place in a.d. 806. There are, therefore, 
no ornamental cross-slabs at Clonmacnois older than 
the beginning of the ninth century. The best ex- 

^ Petrie's Christian Inscriptions in the Irish Language., vol. i., pi. 12, 
Xo, 29. 


amples of recumbent cross-slabs with Celtic ornament 
in Ireland to which reliable dates can be assigned 
are those of Suibine McMailaehumaii at Clonmacnois 
(a.d. 887), and St. Berechtir-^ at Tullylease, Co. Cork. 

Cross-slab from Pen-Arthur, Pembrokeshire 
Now in St. David's Cathedral. Scale l linear 

The latter is specially interesting as having upon 
it a combination of interlaced-work, key-patterns, and 
spiral ornament. 

^ /bid., vol. i., pi. 31, No. 82. - Ib/d., vol. ii., pi. 30. 


Outside the limits of Ireland there are slabs of the 
same type, but of unknown date, at Camborne, ^ Corn- 
wall ; Pen Arthur- (now in St. David's Cathedral), 
Pembrokeshire ; and Baglan,'^ Glamorganshire. 

The recumbent hog-backed or coped stones are more 
likely to be of Anglian or Scandinavian origin than 
Celtic. They are most common in the north of 
England ; there are one or two in Wales, and none in 
Ireland. As instances of coped stones with Celtic 
ornament we have those at Meigle,^ Perthshire ; and 
Lanivet,'' Cornwall. 

The erect cross-slabs are, with a few unimportant 
exceptions, peculiar to Scotland and the Isle of Man. 
They are probably older than the free-standing crosses, 
because the erect cross-slabs are not treated archi- 
tecturally (as the high crosses of Ireland are), but 
resemble more nearly than anything else ornamental 
pages from the Celtic illuminated MSS. directly trans- 
ferred to stone with hardly any modification whatever 
to suit the requirements of the new material to which 
the decoration was applied. A particularly good in- 
stance of this is afforded by the erect cross-slab at 
Nigg,*^ Ross-shire. On one side of the monument is a 
cross with the ornament arranged in rectangular panels 
exactly as it is in the cross-pages of the Irish Gospels ; 
and on the other a figure subject (David and the Lion) 
surrounded by a frame, also divided into panels, as in 
those of the miniatures in the Book of Kells. 

^ Archceologia Camhrensis, sen 5, vol. vi. , p. 357. 
'" Prof. J. O. Westwood's Lapidariurn Wallice^ pi. 60. 
■^ Ibid., pi. 14. 

^ Dr. J. Stuart's Sculptured Stones of Scotland, vol. ii., pi. 131. 
^ A. G. Lang'don's Old Cornish Crosses, p. 412. 

^ Dr. J. Stuart's Sculptured Stones of Scotland, vol. i., p. 28. See 
also casts in the South Kensing-ton and Edinburgh Museums. 

Erect Cross-Slab at Si. IMadoes, Perthshire 
Scale I'.j linear 

1 84 


The following is a list of some of the best specimens 
of erect cross-slabs in Scotland : — 



Papil (now at 



Ulbster (now 

at Thurso) 






) ) 

Hilton of Cadboll (now 

at Inver- 




> > 


) ) 


5 , 




5 } 


. Aberdeensh 


) ) 

The Maiden Stone . 

5 ) 



» » 



Cossins .... 

Farnell .... 

Glamis .... 

Inchbrayock (now at Montrose 


Monifieth (now at Edinburgh) 

St. Vigeans 

Woodwray (now at Abbotsford) 

St. Madoes 

Meigle .... 

Rossie Priory . 









John Patrick of Edinburoh, photo. , , 


Front Back 

Cross at Penmon, AnsJ^lesey 

Drawn by Harold Hughes 

Scale -,\. linear 


The erect cross-slabs of the Isle of Man show a 
mixture of Celtic and Scandinavian art, but there are 
a few which appear to be purely Celtic, as, for instance, 
those at Kirk Maughold^ (on the village green) and 
at Kirk Bride." 

The erect free-standing cross seems to have been 
evolved from the erect cross-slab by removing one 
part of the background of the cross after another, 
until at last nothing but the cross itself was left. We 
see the first stage in the Papil stone from Shetland, 
now in the Edinburgh Museum of Antiquities. Here 
the top of the slab is rounded to suit the curve of the 
circle, within which the head of the cross is enclosed. 
The wheel-cross comes next, in which the portion of 
the background of the cross on each side of the shaft 
is dispensed with, as in the specimens at Margam^ and 
Llantwit Major, ^ both in Glamorganshire. Then the 
ends of the arms of the cross are allowed to project 
beyond the circular ring, as at Penmon,-' Anglesey. 
Lastly, the portions of the background of the cross 
between the quadrants of the ring and the arms are 
pierced right through the slab, thus giving us the 
"four-hole" cross of Cornwall^ and the typical High 
Cross of Ireland. "^ 

We have used the term "wheel-cross" to describe 
the class of monuments with a round head and a shaft 
of less width than the diameter of the head rather 
because it is convenient than on account of its appro- 

^ J. G. Quvam'xn^'s RiDiic Ecmains of t/ie Isle of Mail. 

- Ibid. 

^ Prof. J. O. Westwood's LapidarluDi Wallice, pi. 15. 

^ 76/^., pi. 5. 

^ Ihid.^ pi. 84. 

'^ A. G. Lang^don's Old Covnish Crosses. 

^ H. O'Neill's Ancie)it Crosses of Ireland. 

Great Wheel-Cross of Conbelin at Marg-am Abbey, Glamorg-aiishire 
Drawn by Worthington G. Smith 

Scale I'g linear . "■ 


priateness. Perhaps "disc-cross" would be more 
accurate, but in order to avoid confusion it may be 
as well to adhere to the term " wheel-cross," which has 
been adopted by previous writers on the subject. 

The wheel-crosses are peculiar to Wales, Cornwall, 
and the Isle of Man, there being none either in Ireland 
or Scotland. The wheel-crosses of Wales and the Isle 
of Man have round heads of large diameter and very 
short shafts ; those of Cornwall have heads of much 
smaller diameter with a taller shaft. The best examples 
of wheel-crosses are at Margam and Llantwit Major, 
Glamorganshire ; and at Kirk Braddan and Lonan, 
Isle of Man. 

The free-standing crosses, in which the outline of 
the stone corresponds with the outline of the cross, are 
the most highly developed type of Celtic sculptured 
monument of the Christian period, and are therefore 
presumably the latest, with the exception of those of 
the decadent period just before and after the Norman 
Conquest. The free-standing crosses show the in- 
fluence of the architect rather than that of the monkish 
scribe who embellished the early Irish and Hiberno- 
Saxon illuminated MSS. The sculpture is less flat, 
and the mouldings round the panels of ornament are 
more elaborate than on the earlier erect cross-slabs. 

The free-standing crosses also, instead of being 
monolithic, are constructed of two or more separate 
pieces of stone fixed together by means of mortice and 
tenon joints. In the larger of the High Crosses of 
Ireland the base forms one block, the shaft another, 
the head a third, and sometimes the top arm a 

The High Crosses of Ireland are in most cases 
associated with a characteristic set of ecclesiastical 

Front Back 

Cross at Neuadd Siarman, near Builth, Brecknockshire 

Scale j\ linear 


structures consisting of a Round Tower and several 
small churches. This class of monument consequently 
belongs to the time when the artistic talents of the 
Celtic monks, which had been previously entirely 
absorbed in illuminating MSS., was directed into 
the new channel of architecture. The High Cross 
of Muiredach^ at Monasterboice, Co. Louth, and that 
of King Fland' at Clonmacnois, King's Co., are 
proved by the inscriptions upon them to have been 
erected during the first quarter of the tenth century. 
There is such a general family likeness between most 
of the High Crosses of Ireland that they are probably 
all of about the same date. 

There is a peculiarity in the design of some of the 
High Crosses of Ireland which should not pass un- 
noticed, namely, the semicircular projection in each of 
the four hollows between the arms.^ In a stone cross 
these projections have no use or meaning, but in the 
metal crosses of the same period projections of this 
kind serve to diguise the rivets by means of which the 
metal plates on each side of the cross are held together.* 
From this it would appear that the art of the worker in 
metal to some extent influenced the sculptors by whom 
the stone crosses were made. 

Some of the Cornish crosses have triangular pro- 
jections in a similar position, giving an appearance not 
unlike the cusping in Gothic window tracery. 

The free-standing crosses of Wales and Cornwall 
differ from those of Ireland in having heads of much 

^ Petrie's Christian Inscriptions in the Irish langtiage, vol. ii., p. 66. 

" Ibid.^ vol. i., p. 43. 

^ As on the crosses at Monasterboice, Co. Louth, and Durrow, 
King's Co. 

^ As on the Cross of Cong- in the Dublin Museum, and on the pectoral 
cross of St. Cuthbert in the Library of Durham Cathedral. 



M^-n g 










Cross at Nevern, Pembrokeshu'e, ^ j' 
Scale Tj'j linear . " • ^ > ,^ j 


^ F 



smaller diameter in proportion to the height of the 
shaft, and bases are the exception rather than the rule. 
In the Welsh and Cornish crosses figure sculpture is 
made altogether subordinate to ornament, whilst in the 
Irish crosses exactly the reverse is the case. The 
fronts and backs of the Irish crosses, and sometimes 
the sides also, are entirely covered with panels of 
symbolical figure subjects forming a cycle, which does 
not occur in the illuminated MSS., although evidently 
borrowed from a Byzantine source. The subordination 
of ornament to figure subjects on the Irish crosses 
shows that they are further removed from the MSS. 
than the Welsh, Cornish, and Scottish crosses, and 
therefore of later date. The free-standing crosses of 
Scotland seem to belong to the Irish group. 

The following list gives the best examples of free- 
standing crosses : — 


Kells . 









Clonmacnois . 


g's Co. 



g's Co. 

Castle Dermot 



Moone Abbey 



Kilklispeen . 




















. Perthshire. 



Penmen Anglesey. 

Maen Achyfan 


Neuadd Siarman 


Llanbadarn Fawr 


Llantwit Major 



5 » 




) » 


y » 

The shafts of the erect free-standing crosses which 
have just been described are rectangular in section, 
but there are a few exceptional monuments with shafts 
of square section or of round section, or partly of 
square and partly of round section. As an instance 
of a cross of square section we have the one at 
Llandough, Glamorganshire. At Llantwit Major, in 
the same county, is a cylindrical pillar with a vertical 
groove down one side of it, the use of which has 
caused much futile speculation amongst antiquaries. 
The pillar of Eliseg at Valle Crucis, Denbighshire, is 
round at the bottom and square at the top, thus corre- 
sponding in shape to a well-known type of monument 
which is common in Mercia. These round pillar 
crosses usually occur in pairs. 

There are a few unique monuments that cannot be 
classed with any of those already described, such as 
the ornamented stone coffin at Govan, Renfrewshire, 
and the altar tomb at St. Andrews, Fifeshire. 

Descriptions and illustrations of nearly all the 
monuments mentioned will be found in Dr. J. Stuart's 
Sculptured Stones of Scotlmid (published by the 
Spalding Club of Aberdeen) ; Dr. J. Anderson and 
J. R. Allen's Early Christian Monuments of Scotland 


(published by the Society of Antiquaries of Scotland) ; 
R. C. Graham's Carved Stones of Islay; H. O'Neill's 
Sculptured Crosses of Ancient Ireland; Dr. Petrie's 
Christian Inscriptions in the Irish Language (published 
by the R. Soc. Ant. of Ireland) ; Miss M. Stokes' 
Early Christian Art in Ireland; Prof. J. O. West- 
wood's Lapidarium Wallicc ; A. G. Langdon's Old 
Cornish Crosses ; J. G. Cumming's Rimic Remains of 
the Isle of Man. 

The Celtic metalwork of the Christian period may be 
arranged under the following heads : — 


Processional crosses. 
Bell shrines. 

Book shrines. 
Relic shrines. 
Plaques for book-covers. 
Penannular brooches. 
Hammer-headed pins. 

With a few exceptions all the existing specimens are 
now preserved in the Museum of the Royal Irish 
Academy in Dublin, the National Museum of Anti- 
quities of Scotland in Edinburgh, and the British 
Museum in London. 

Ecclesiastical bells are of two different kinds, namely, 
{i) portable bells, sufficiently light to be carried in the 
hand ; and (2) fixed bells, whose weight renders a 
trussed framework of wood necessary for their support. 
Each kind of bell can be rung in two separate ways, 
namely, (i) by holding the bell stationary and striking 
it on the outside with a hammer ; or (2) by providing 
the bell with a tongue, or clapper, suspended from the 
inside and swinging the bell backwards and forwards, 
so as to cause the clapper to strike against the interior 
and thus produce sound. The method of bell-ringing 
by means of a hammer is frequently illustrated in the 


illuminated psalters of the thirteenth and fourteenth 
centuries, and is also to be seen on the sculptured 
capitals in the Abbey of St. George's de Boscherville,^ 
in Normandy. The great bells of the Kremlin at 
Moscow, and in other Greek churches throughout 
Russia, are rung in this fashion. Portable bells with 
clappers have a handle at the top, by which they can 
be swung backwards and forwards in the hand, in the 
manner depicted upon the Bayeux Tapestry." Fixed 
bells with clappers have loops at the top for suspension 
by iron bands to a horizontal wooden axle or rocking 
bar working in bearings supported on a trussed frame- 
work of timber, usually within a masonry tower. The 
required rocking motion is given by a lever and rope 
or a grooved wheel and rope. 

The bells used in the Celtic Church seem to have 
belonged exclusively to the class of portable bells rung 
by hand. During the earlier period of Christianity in 
Ireland, when the monks lived together in small isolated 
communities, bells which were intended to carry sound 
to a great distance would be unnecessary, so that the 
absence of belfries in connection with the primitive 
dry-built stone oratories of the sixth and seventh 
centuries is easily explained. When, however, at a 
later period, the congregations became larger and 
more widely scattered, the lofty tower served a useful 
purpose in greatly increasing the area over which the 
sound of the bell could be heard. 

The commencement of the building of belfries in 
Ireland coincides with the introduction of Lombardo- 
Byzantine architecture into that country, and the Irish 
round tower is obviously nothing more than a local 

^ Didron's Amiales ArcJiMogiqiies^ vol. vi., p. 315. 
^ F. R. Fowke's Bayeux Tapestry^ pi. 31. 


variety of the Italian campanile. The Viking in- 
vasions at the same time gave an additional impetus to 
the erection of structures which could be used not only 
for ecclesiatical purposes, but also as watch-towers to 
detect the approach of the enemy, as bell-towers to 
alarm the neighbourhood, and as towers of defence 
to secure the lives and property of the congregation. 
The fact that the Irish round towers are called by the 
name of cloiccthec, or bell-house, in the ancient annals 
is sufficient proof they were used as belfries, but it 
does not appear to be known whether the bells were 
rung by swinging in the hand or fixed to a framework 
and swung on pivots. At any rate, no Irish bells of 
this period (a.d. 800 to 1000) have survived except the 
portable hand-bells. If any mechanical appliance was 
employed for bell-ringing in the Irish round towers it 
was probably constructed by fixing an ordinary hand- 
bell to a horizontal axle-bar of wood or iron, working 
in two bearings, and swung backwards and forwards 
by means of a rocking lever with a rope attached to it, 
as is done in many village churches at the present day. 
The large, heavy metal bells made specially with a 
view to being fixed in a tower and rung by a grooved 
wheel and cord belong to a much later period, after the 
Norman Conquest, when the art of making castings in 
bronze of great size had been learnt. 

The portable bell of the early Celtic Church is 
merely an ordinary cattle bell,^ such as would, no 
doubt, be common in Pagan times, adapted to eccle- 
siastical purposes and slightly modified to suit the 
requirements of the monks. It differs hardly at all, 
except as regards size, from the common sheep-bell 

^ Probably the earliest representation of a cow-bell in Great Britain 
is on the pre-Norman cross at Fowlis Wester, near Crieff, Perthshire, 


still to be found in many parts of England. Dr. 
Joseph Anderson tersely sums up the peculiarities of 
the Celtic ecclesiastical bell, as regards its material, 
manufacture, form, and size, in his Scotland in Early 
Christian Times (first series), p. 183, somewhat as 
follows : — 

(i) Material — iron coated with bronze. 

(2) Manufacture — hammered and riveted ; coating of 

bronze put on by means of a process analogous 
to tinning. 

(3) Form — tall, narrow, tapering, foursided ; ends 

flattened ; sides bulged. 

(4) Size — portable ; provided with handle so as to be 

easily swung by hand. 

The original home of ecclesiastical bells of this type 
was in Ireland, where there are still the greatest num- 
ber in existence, and thence they spread to Scotland, 
Wales, England, Brittany, France, and Switzerland. 

The largest iron bell of this kind is preserved in 
the Church of Birnie, near Elgin, N.B. It is i foot 
2 inches high, and 7 inches by 5 inches at the bottom, 
tapering to 4J inches by 3 inches at the top. It is 
riveted down each of the narrow sides with four rivets, 
and the handle is fixed to the top by four much smaller 
rivets. As a rule, however, the height of such bells 
rarely exceeds i foot or is less than 8 inches. 

The Celtic ecclesiastical bell of wrought-iron was 
afterwards copied in cast bronze. It is reasonable to 
suppose that the bronze bells are of later date than those 
of iron (i) because the rectangular shape is useless and 
meaningless in the case of a bronze bell, and results 
from copying an iron bell, in which the rectangular 
shape is necessitated by its method of construction ; 
(2) because the bronze bells are of more refined shape 


and better manufacture than those of iron ; and (3) 
because the bronze bells are in many cases ornamented. 
Celtic ecclesiastical bells of cast bronze may be 
divided into the following classes : — 

(i) Bronze bells without ornament. 

(2) Bronze bells without ornament, but inscribed. 

(3) Bronze bells with ornamented handles. 

(4) Bronze bells with ornamented bodies. 

Examples of Celtic quadrangular bells of cast bronze 
without ornament have been recorded at the following 
places : — 

Wales — 

Llanrhyddlad, Anglesey {ArchcEologia Ccunbrensis, 4th 

sen, vol. ii., p. 275). 
Llangystenyn, Carnarvonshire ; now in the Powysland 

Museum at Welshpool {Mojitgomeryshire Collections, 

vol. XXV., p. 327). 

Scotland — 

Eilean Finan, Loch Shiel, Argyllshire (Dr. J. Anderson's 

Scotland in Early Clwistian Times, ist ser. , p. 198). 
Insh, near Kingussie, Inverness-shire {Ibid., p. 195). 
Little Dunkeld, Perthshire (Proc. Soc. Ant. Scot., vol. 

xxiii., p. 119). 
Forteviot, Perthshire {Ibid., vol. xxvi., p. 434). 

Ireland — 

Garton, Co. Donegal (Rev. H. T. Ellacombe's Church 

Bells of Devon — Supplement, p. 342). 
Lower Badony, Co. Tyrone (Ibid., p. 344). 
Scattery Island, Co. Clare ; now in the British Museum 

(Ibid., p. 344). 
Kilbroney, Rostrevor, Co. Down (R. Welch, photo. 

No. 1,932; Jour. R. Soc. Ant. of Ireland, vol. xxxiii., 

P- 55)- 
Kilmainham {Jour. R. Soc. Ant. of Ireland, 5th ser., 

vol. x. , p. 41). 


Ireland — 

Cappagh, Co. Tyrone (Jour. R. Soc. Ant. of Ireland, 

vol. xxxiii., p. 52). 
Drumragfh, Co. Tyrone {Ibid.^ vol. xxxiii., p. 54). 


Goulien, Finistere {Ibid., 5th ser. , vol. viii., p. 167). 

As has already been pointed out, the bells of cast 
bronze are copies in another material of the wrought- 
iron bells, the quadrangular form of which had its origin 
in the method of construction out of a thin sheet of 
metal with riveted joints being still adhered to in the 
bronze bell, where joints were not required. The only 
difference in the shape of the iron and the bronze bells 
is that the latter have in most cases a flange, or an 
expansion and thickening of the metal round the 
mouth. The handles vary from those which are 
almost rectangular to those which are quite round. 
The bell still preserved in the church at Insh, near 
Kingussie, Inverness-shire, may be taken as a fair 
sample of the Celtic quadrangular bell of cast bronze 
without ornament. It is 10 inches high, and measures 
9 inches by 7f inches at the mouth. The handle is 
oval and the mouth expanded. The remaining bells 
of the same class vary from 4 inches to 11 inches in 
height, with their other dimensions in proportion. 

There are three Celtic quadrangular bells of cast 
bronze without ornament, but inscribed, at the follow- 
ing places : — 

Ireland — 

Clogher, Co. Tyrone (H. T. Ellacombe's Church Bells 

of Devon— Supplement, p. 369). 
Armagh ; now in the Museum of the R.I. A. at Dublin 

(M. Stokes' Early Christian Art in Ireland, p. 65). 


Brittany — 

Stival {Memoii'es de I'Institut hnperiale de France; 
Academie des Inscriptions et Belles-Lettres^ vol. xxiv. , 
pt. ii., p. 387). 

The bell of Clogher is inscribed, in one horizontal 
line, with Roman capital letters — 


The bell of Armagh is inscribed, in three horizontal 
lines, with Hiberno-Saxon minuscules — 

►J< oroit ar chu 

mascach m 


" >J< A prayer for Cumascach, son of Ailell." 

The bell of Stival is inscribed, in one vertical line, 
with Carlovingian minuscules — 

pirtur ficifti 
"Pirtur made this" (?). 

Or, according to the Vicomte Hersart de la Villemarquc : 

pir turfic is ti 
" Sweet-sounding" art thou." 

The Cumascach mentioned on the bell of Armagh 
was probably the steward of Armagh, who, according 
to the Annals of the Four Masters^ died in a.d. 904, 
thus fixing the date of at least one of the bells of this 

Celtic quadrangular bells of cast bronze with orna- 
mented handles exist at the following places : — 

Wales — 

Llangwynodl, Carnarvonshire ; now in the possession 
of W. C. Yale-Jones-Parry, Esq., of Madryn Castle, 
Pwllheli, Carnarvonshire {Archceologia Carnbrensis^ 
ist ser. , vol. iv., p. 167 ; and 4th ser. , vol. ii., p. 274). 


Scotland — 

Strathfillan (Bell of St. Fillan), Perthshire ; now in the 
National Museum at Edinburgh (Dr. J. Anderson's 
Scotland in Early Christian Times ^ ist ser. , p. i86). 
Ireland — 

Lorrha (Bell of St. Ruadhan), Co. Tipperary ; now in 
the British Museum (H. T. Ellacombe's Church Bells 
of Devon — Supplenieiit^ p. 344). 
France — 

St. Pol de Leon (Bell of St. Meriadec) (Rohault de Fleury's 
La Messe, vol. vi., pi. CDXvin. ; Ellacombe, p. 383). 

The ornament on the handles is of two kinds — 
zoomorphic and phyllomorphic. The former consists 
of the head of a beast at each end of the loop handle 
where it joins the body of the bell, and the latter of a 
leaf in the same position. The bell of Llangwynodl^ 
has a good typical example of a zoomorphic handle, 
and the bell of St. Pol de Leon is the only one with 
leaf terminations to the handle. The Llangwynodl 
bell is 5 inches high, and measures 6\ inches by 
4 inches across the mouth ; and the St. Pol de Leon 
bell is gh inches high, and measures 6h inches across 
the mouth. St. Fillan's bell is i foot high, and 
St. Ruadhan's bell only 2 inches or 3 inches high. 

Celtic quadrangular bells of cast bronze with orna- 
mented bodies exist at the following places : — 

Ireland — 

Lough Lene Castle, Co. Westmeath ; now in the Museum 
of the R.I.A. at Dublin; Bangor, Co. Down {Ulster 
Journal of Arc hcBO logy, vol. i., p. 179 ; Ellacombe, p. 340). 
Cashel ; now at Adare Manor (Lady Dunraven's Mem- 
orials of Adare Manor ^ p. 152 ; Ellacombe, p. 340). 

^ We are indebted to Mr. W. Corbet Yale-Jones-Parry, of Madryn 
Castle, Pwllheli, the present owner of the Bell, for permission to re- 
produce the photograph. 


By the courtesy of Mr. George Coffey, m.r.i.a., of 
the Museum of the Royal Irish Academy, we are 
able to illustrate the bell from Lough Lene Castle. 

It is I foot ij inches high, including the handle, 
and measures 8^ inches by 7} inches across the mouth. 
The shape of the body of the bell resembles that of 
the iron quadrangular bells, but exhibits much greater 
refinement in the delicate and almost imperceptible 
curves of the sides. The handle is semicircular. The 
cross of the well-known Irish type, with a border of 
key pattern below, round the mouth of the bell, on 
one of the border faces ; and a border of angular 
interlaced-work in a similar position on each of the 
narrower faces. 

The bell of Bangor was found at the place of that 
name, in Co. Louth, and was subsequently in the 
possession of Dr. Stephenson, of Belfast. It now 
belongs to Colonel MacCance, of Knocknagoney 
House, Holywood, Co. Down.^ It is i foot 2h inches 
high, and measures 9 inches by 8 inches across the 

This bell is also ornamented with a cross and key 
patterns, like the one just described, the only differ- 
ence being that the cross is not combined with a 
circular ring, and the design of the key pattern is not 
quite the same. 

The bell of Cashel was found at the place of that 
name, in Co. Tipperary, in 1849, and is now preserved 
at Lord Dunraven's house at Adare Manor, Co. 
Limerick. It resembles the bell of Bangor almost 
exactly, except that there are four round dots in the 

^ Mr. R. Welch, of Belfast, tells me that it is kept in a hre-proof 
safe, and that over jCs^o was refused for it. 




:> .y^':> 


hollows between the arms of the cross. The handle is 
broken off, and without this the bell is i foot high. 
Its dimensions across the mouth are 9J inches by 
6i inches. 

These three bells are so nearly alike as regards their 
size, shape, and ornamentation that they are probably 
all the same date, and may even have been the work of 
one artificer in metal. A peculiarity occurs in the key 
patterns on the bells from Lough Lene Castle and from 
Bangor which may perhaps help to fix the date. It 
will be noticed that the square spaces in the middle of 
the key patterns are filled in with an almond-shaped 
figure. This is also a feature of the key patterns in 
the Irish Gospels (Codex No. 51) at St. Gall, in 

There is in the British Museum a Celtic quadrangular 
bell of iron with an ornamental bronze cap fixed to the 
top of it, but it is not clear whether the cap forms part 
of the original design or was added subsequently. 
This bell is called the Bell of Conall Gael, and came 
from Inishkeel, in the Barony of Boylagh, Co. 
Donegal. It was enclosed within a metal shrine in the 
fifteenth century. 

All the other Celtic ecclesiastical bells which have 
been enshrined are entirely of iron, a fact tending to 
show that the bronze bells are of later date than the 
iron ones, because the enshrined bells were those 
belonging as a general rule to the saint who founded 
the church. The bronze bells probably came into 
use long after most of the older churches had been 

^ R. Purton Cooper's Appendix A to Rymer's Fcedera, p. 90 and 
pi. 7 (St. Mark miniature), and pi. 10 ^'initial pag-e of St. John's Gospel). 


It may be interesting to give a list of the bell-shrines 
still in existence : — 

Ireland — 

Shrine of the Bell of St. Patrick's Will ; now in the 

Museum of the R.I.A. at DubHn (H. O'Neih's Fine 

Arts of Ancient Ireland^ p. 46). 
Shrine of the Bell of St. Culan, called the Barnaan 

Ciiilaun ; now in the British Museum {Transactions of 

the Royat Irisli Academy^ vol. xiv. , p. 31). 
Shrine of the Bell of St. Mogue. 
Shrine of the Bell of Maelbrigde (Miss M. Stokes' 

Early Christian Art in Ireland^ p. 67). 
Shrine of the Bell of St. Mura, from the Abbey of 

Fahan, Co. Donegal {Ulster foiirnal of A^'chcEology ^ 

vol. i., p. 274). 
Shrine of the Bell of Conall Gael, from Inishkeel, 

Co. Donegal ; now in the British Museum (H. T. 

EUacombe's Church Bells of Devon — Supplenieiit^ 

P- 365)- 
Scotland — 

Bell-shrine of Kilmichael Glassary, Argyllshire, dug up 

on Torrebhlaurn Farm in 1814 ; now in the National 

Museum at Edinburgh (Dr. J. Anderson's Scotla7id in 

Early Christian Tinies^ ist ser. , p. 207). 
Bell-shrine, preserved at Guthrie Castle, Forfarshire 

{Ibid., p. 209). 

The bells of the Celtic Church, whether they be of 
iron or bronze, whether devoid of lettering or in- 
scribed, ornamented or plain, possess a far higher 
interest than that attaching to ordinary museum speci- 
mens, because most of them have an authentic history, 
going back in some cases to the time when Christianity 
was first introduced into this country. The bell, the 
book, and the crozier which belonged to the Celtic 
saints who founded churches, were always looked upon 


A.D. I091 TO I 105 


with the highest veneration, and were used for a variety 
of superstitious purposes, such as healing the sick, pro- 
curing victory in battle, and the solemnising of oaths. 
The relics of the saints of the fifth, sixth, and seventh 
centuries were enclosed in costly metal shrines, gene- 
rally a few hundred years after the death of the 
saint, and an hereditary keeper was appointed to be 
responsible for the safety of the relics when borrowed 
for effecting cures and other purposes. The shrines 
and their contents were thus handed down from gene- 
ration to generation, and in most cases sold by their 
last hereditary keepers to collectors of antiquities, from 
whom they were acquired by the national museums of 
England, Scotland, and Ireland. The relics still bear 
the names of the saints to whom they originally be- 
longed ; the names of their hereditary keepers are well 
known, and they have been obtained from the localities 
where the saint founded his church, and where the 
relics remained for centuries afterwards undisturbed. 
No class of antiquities, therefore, possesses a better 
record or a more satisfactory pedigree. 

The Irish and Scottish bell-shrines which have been 
enumerated are cases of metal of the same shape as 
the bell they contain, having four sloping sides and an 
arched top. The sides are usually made of bronze 
plates ornamented with gold, silver, enamel, and set- 
tings of crystal and precious stones. Two features 
which are characteristic of the ornamental bronze bells 
are repeated in the shrines, namely, the zoomorphic 
terminations of the handles and the cross on the body 
of the bell. In the two Scottish bell -shrines the 
Crucifixion takes the place of the Cross. The orna- 
ment on the bell-shrines is much further removed in 
style from that of the illuminated MSS. than is the 


case with the sculptured stones. This is only what 
might be expected, considering the late date of the bell- 
shrines as compared with that of the crosses. On two 
of the bell-shrines Scandinavian influence may be 
clearly detected in the ornament upon them. Thus on 
the Shrine of the Bell of St. Patrick's Will the pear- 
shaped eyes of the beast's heads on each side of the 
arched top are placed with the point outwards in the 
Scandinavian fashion ; and on the Shrine of the Bell 
of St. Mura the ''tendril pattern," which is so common 
on the Rune-inscribed monuments of the Isle of Man, 
may be noticed. 

The dates of three of the bell-shrines have been 
ascertained by means of the inscriptions upon them, 
namely, Maelbrigde's Bell-shrine, circa a.d. 954; the 
Shrine of the Bell of St. Patrick's Will, a.d. 1091 
to 1 105; and the Guthrie Bell-shrine, 14th century. 
Judging merely from the style of the ornament, the 
Shrine of the Bell of St. Culanus should be of the 
twelfth century, and the Shrines of the Bell of St. 
Mura and of Kilmichael Glassary perhaps as late as 
the beginning of the thirteenth century. 

The metal croziers of the Celtic Church are in reality 
shrines enclosing the wooden pastoral staffs of the 
difl'erent saints, whose names most of them still bear. 
The chief peculiarity of the Celtic crozier is the shape 
of the head, which is like the hook of a modern 
walking-stick, but with a remarkable flattened end. 
The inside curve of the hook is nearly circular, but 
the outside curve is only partially semicircular, and 
suddenly changes to a nearly vertical straight line 
just before the end of the crook is reached. At the 
bottom of the crozier is a pointed ferrule, and the 
straight portion consists of two cylindrical tubes of 


A. D. 1090 TO 1 1 13 


thin metal joined together in the middle by a bulbous 
collar. The upper tube is joined to the head at the top 
by a similar bulbous collar, and the lower tube is joined 
to the ferrule at the bottom by a third bulbous collar. 

One of the most perfect of the Irish crozers is pre- 
served at Lismore Castle,^ Co. Waterford. It bears an 
inscription showing that it was made by Nectan, the 
artisan, for Niall, son of MacAeducain. Mac Mic 
Aeducain was Bishop of Lismore from a.d. 1090 to 
1 1 13. Another fine crozier in the British Museum- 
has an inscription asking a prayer for Maelfinnia and 
Condulig. The former was Bishop of Kells, and died 
in A.D. 967. Condulig was an ecclesiastic of the same 
monastery, and died in a.d. 1047. 

The best examples of uninscribed croziers are the 
croziers of Clonmacnois'^' and of St. Berach in the 
Museum of the Royal Irish Academy in Dublin, and a 
crozier now in the possession of the Roman Catholic 
Bishop of Killarney. Besides the complete croziers 
mentioned there are several heads and other portions of 
croziers in the Museum of the Royal Irish Academy, 
and in the Edinburgh Museum of Antiquities may be 
seen the head of the crozier of St. Fillan,'^ which has 
an unusually interesting history. 

The decoration of the Celtic croziers is concentrated 
on the head, the ferrule, and the collars round the 
straight portion of the staff. Most of the croziers 
have a zoomorphic cresting'"* on the outside curve of 
the head, sometimes consisting of a procession of 

^ Petrie's Christian Tnscriptio?is in the Irish Language, vol. ii., p. ii8; 
and H. O'Neill's Fine Arts of Ancient Irelayid, p. 42. 

2 Petrie's Christian Inscriptions in the Irish Language, vol. ii., p. 116. 
^ Miss M. Stokes' Early Christian Art in Ireland, p. 105. 
'^ Dr. J. Anderson's Scotland in Early Christian Times, ist sen, p. 219, 
^ As on the croziers of Lismore, Clonmacnois, and Dysert. 


beasts one behind the other, and sometimes only 
having terminal beasts' heads at each end.^ The flat 
portion of the crook of the crozier at the end is 
decorated in some cases with the head of the saint 
or bishop, and a crystal setting below.- Zoomorphism 
enters very largely into the ornamentation of the Celtic 
croziers, and the beasts with only two toes instead of 
three on the Crozier of Clonmacnois obviously betray 
their Scandinavian origin by this detail. The decora- 
tion of the heads of the croziers is treated in at least 
three different ways: (i) the head of the Lismore 
crozier is divided into rectangular panels with raised 
bosses of enamel at the intersections of the bands, 
which form the divisions between the panels ; (2) the 
heads of the croziers of Dysert, Blathmac, and St. 
Fillan are divided into lozenge-shaped panels by a 
sort of raised lattice-work ; and (3) the head of the 
crozier of Clonmacnois is not divided into panels, but 
the surface entirely covered with zoomorphic strap- 

The croziers are all of the eleventh century or later, 
and their decoration has little in common with that of 
the early illuminated MSS. 

*'Cumdachs,"or book-shrines, are peculiar to Ireland. 
Three MSS. still in existence are known, from historical 
evidence, to have had cumdachs, although they have 
been lost. 

These are : — 

The Book of Durrow enshrined a.d. 877 to 914. 

The Book of Armagh ,, a.d. 938. 

The Book of Kells ,, before a.d. 1007. 

^ As on St. Fillan's crozier. 
■^ As on St. Fillan's crozier, 


The existing cumdachs are as follows : — 

In the Museum of the Royal Irish Academy. 
Cumdach of Molalse's Gospels . a.d. iooi to 1025. 
,, ,, the Stowe Missal . a.d. 1023. 

,, ,, Columba's Psalter . a.d. 1084. 

,, ,, St. Patrick's Gospels. 

In the Library of Trinity College^ Duhli^i. 
Cumdach of Dimma's Book . a.d. 1150. 

The cumdachs are simply rectangular boxes, suffi- 
ciently large to hold the MS., made either of wood or 
bronze and plated with silver. The decoration of the 
principal face of the cumdach is generally arranged in 
the form of a cross, the treatment being much the 
same as in the ornamental pages of the MSS. of the 
Gospels. The cross on the cumdach of Molaise's 
Gospels is formed of a flat silver plate with panels 
pierced right through the thickness of the metal, and 
filled in with interlaced patterns in filigree-work. The 
cross in the middle is surrounded by the Symbols of 
the Four Evangelists, with their names inscribed at 
the side of each. The centre of the cross and the 
ends of the four arms are ornamented w^ith settings of 

On one of the narrow faces of this cumdach are 
some very curious figures of two ecclesiastics, one 
holding a bell and the other a pastoral staff; and a 
harper, with an angel above his head, between them. 

The cumdach of the Stowe Missal has upon the 
principal face a cross within a rectangular frame. The 
centre of the cross is ornamented with a crystal setting, 
and the recessed panels of the background are filled in 
with a peculiar kind of triangular and square chequer- 
work made of pierced metal plates. 


The cumdach of Dimma's Book has also on the 
principal face a cross surrounded by a rectangular 
frame, and ornamented with thirteen crystal settings. 
The four recessed panels of the background of the 
cross are filled in with zoomorphic designs in the same 
style as those on the High Cross of Tuam,^ Co. 
Galway, which is of about the same period, having 
been erected in a.d. 1123. 

The relic-shrines of the Celtic Church are of two 
kinds, namely, (i) those made in the shape of the 
portion of the body of the saint enshrined ; and (2) 
those made in the shape of a small oratory or house 
with a steep pitched roof having hipped ends. As 
an example of the first kind we have the Shrine of 
St. Lachtin's Arm."^ 

The most beautiful and perfect example of a reli- 
quary in the form of a small oratory is the one now 
in the possession of Sir Archibald Grant, and preserved 
at Monymusk House, -^ Aberdeenshire. It is a wooden 
box, hollowed out of the solid, and covered with plates 
of bronze and silver. It is decorated with enamel, 
settings of precious stones, and raised circular medal- 
lions and rectangular plaques of interlaced-work on a 
chased background of zoomorphic designs. Another 
reliquary of the same kind was found in Lough Erne,^ 
between Enniskillen and Belleek, in 1891, and belongs 
to Mr. T. Plunkett, of Enniskillen. It is 7 inches 
long by 3J inches wide by 5!^ inches high, and is made 
of plates of bronze covering an inner box scooped out 
of two solid pieces of yew-wood. The decoration, 

' H. O'Neill's Sculptured Crosses of Ancient Ireland^ pi. 12. 

- Vctusta Monunienta^ vol. vi., pi. 19. 

" Dr. J. Anderson's Scotland in Early Christian Times, rst ser. , p. 249. 

** Journ. R. Soc. Ant. of Ireland, 5th ser., vol. ii. (1892), p. 349. 





ir. Morgan Evam;, of Pwlllieli, plioto. 



R. ll'ekh, of Belfast, /■hoto. 


which consists of interlaced-work, is concentrated upon 
the ridge-piece of the roof; upon a band concealing 
the joint between the eaves of the roof and the sides ; 
upon six circular raised medallions, one on each of the 
longest sloping faces of the roof and two on each of 
the longest sides ; and upon the hinges at each end 
of the box to which the bars for suspending the shrine 
round the neck of its hereditary keeper were attached. 
There is a third reliquary, like the two just described, 
from Norway, in the Copenhagen Museum.^ It has 
raised circular medallions arranged in the same way as 
on the Lough Erne shrine, but they are decorated with 
spiral designs, and the background, instead of being 
plain, is covered with elaborate interlaced-work. An 
inscription in later Runes on this shrine reads ''Ran- 
vaig owns this casket." The Edinburgh Museum 
possesses a fourth shrine of the same class found in 
the Shannon^ in a very dilapidated condition. 

Dr. J. Anderson-^ has pointed out the identity of the 
form of the Temple at Jerusalem, as represented in the 
Book of Kells, with the form of this particular class of 

The Breac Moedoc,^ or shrine of St. Mogue, from 
Drumlane, now in the Museum of the Royal Irish 
Academy at Dublin, resembles the reliquaries of the 
Monymusk type in shape except that the roof is gabled 
instead of being hipped, and the method of applying 
the decoration also is entirely different. It is 7J inches 
long by 8f inches wide by si inches wide, and is made 
of bronze, with decorations of bronze gilt, enamel, 

^ J. J. A. Worsaae's Nordiske Oldsager i det Kongelige Museum i 
Kjdbe7ihavn, p. 129, Fig-. 524. 

'■^ Dr. J. Anderson's Scotland in Early Christian Tifties^ 1st ser. , p. 246, 

'^ Ihid., p. 247. 

^ Archceologia, vol. xliii., p. 131. 


and glass. The front is divided into rectangular 
panels, each containing a group of figures of male and 
female saints numbering twenty-one altogether ; and 
on one of the gabled ends is a bearded figure playing 
a harp on which a bird is perched. The back and 
bottom of the shrine are ornamented with cruciform 
patterns in pierced work, as on the shrines of the Bell 
of St. Patrick's Will, of the Stowe Missal, and of 
Dimma's Book. 

The relic-shrine of St. Manchan^ differs from all 
those previously described in being considerably larger, 
and in being shaped like the gabled roof of a house, 
but without any house ; that is to say, it has two rect- 
angular faces meeting in a horizontal ridge and 
two nearly vertical triangular ends. It was for- 
merly in the keeping of the ancient Irish family of 
Mooney, of the Doon, but it is now preserved in 
the Roman Catholic Church of Boher, in the parish 
of Lemanaghan, near Clara, King's Co. The Shrine 
of St. Manchan is i foot ii inches long by i foot 
I inch wide by i foot 7 inches high. The fram.e- 
work of the shrine is made of yew boards. The front 
and back are each ornamented with an equal-armed 
cross having large circular raised bosses in the centre, 
and on the ends of the four arms. The four spaces 
forming the background of each of the crosses are 
filled in with rows of small figures fixed to the bronze 
plate behind with rivets. The front, back, and two 
ends of the shrine are partially surrounded by a border 
of zoomorphic ornament. The bosses in relief of the 
crosses on the front and back, and the recessed 
triangular panels on the two ends are also elaborately 
decorated with zoomorphs. At each of the four corners 

^ The Reliquary^ vol. xv. (1875), p. 193. 


of the base is a circular ring, probably for carrying the 
shrine about. The clamps of the rings, the borders 
round the bottom of the shrine, and the narrow parts 
of the arms of the crosses have step-patterns in red 
and yellow enamel upon them. The whole of the 
bronze was originally gilt. The style of the ornament 
is so similar to that on the Cross of Cong that we shall 
not be far wrong if we attribute the shrine of 
St. Manchan to the same period, namely, the twelfth 

There is only a single example of a processional 
cross belonging to the Celtic Church now in existence, 
namely, the Cross of Cong^ in the Museum of the 
Royal Irish Academy at Dublin. It is 2 feet 6 inches 
high by i foot 6f inches across the arms by ij inches 
thick. The cross is of oak covered with copper plates, 
and has a boss of rock-crystal in the centre, beneath 
which the portion of the true cross was enshrined. 
The outer margin of the cross is formed by a roll 
moulding of silver, with eighteen small enamelled 
knobs at intervals to emphasise the cuspings of the 
outline of the cross. The face of the cross within the 
margin is divided into two rows of panels by a narrow 
longitudinal band in the middle of the arms, with 
enamelled bosses of enamel in relief and circular silver 
discs alternately marking the points where the cross- 
bars branch off at right angles to the central stem, so 
as to divide the surface into panels. The eight panels 
surrounding the boss of rock-crystal in the centre of 
the cross are filled in with scrolls of gold filigree-work, 

^ Proc. R.I. A., vol. ii., p. 113, and vol. iv. , p. 572 ; Petrle's Christian 
Inscriptions in the Irish La?iguage, vol. ii., p. 118; Miss M. Stokes' 
Early Christian Art in Ireland^ p. 108 ; and /our n. R. Soc, Ant. Ireland^ 
vol. xxxi. (1901), p. 40. 


and the remaining thirty-eight panels on the arms 
and shaft are filled in with zoomorphic designs in cast 
bronze gilt, riveted to the copper plates beneath. At 
the bottom of the cross is a beast's head with a bulbous 
projection between it and the socket to receive the staff. 
The bulbous portion is ornamented with small bosses 
of blue enamel and panels of zoomorphic designs. 
The general effect of the whole is extremely rich, and 
shows great artistic feeling. The prevalence of the 
zoomorphic element in the design and the arrangement 
of the panels reminds us of the croziers of the same 
period, more especially the one at Lismore Castle, 
Co. Waterford. 

The inscriptions on the Cross of Cong, of which the 
first is in Latin (twice repeated) and the remaining four 
in Irish, may be thus rendered in English : — 

(i) "This Cross covers the Cross on which the Saviour 
of the World suffered." 

(2) " Pray for Murdoch O'Duffy, the Senior of Ireland." 

(3) "Pray for Turloch O'Connor, for the King of Ire- 

land, for whom this shrine was made." 

(4) "Pray for Donnell M'Flannagan O'Duffy, for the 

Bishop of Connaught, for the successor of Coman 
and Ciaran, under whose superintendence this 
shrine was made." 

(5) "Pray for Maeljesu MacBratdan O'Echan, who 

made this shrine." 

Murdoch O'Duffy, Archbishop of Connaught, died 
in A.D. 1 150, and it is recorded in the Annals of Tnnis- 
fallen that in the year 11 23 a bit of the true cross came 
into Ireland and was enshrined by Turlogh O'Connor, 
thus fixing the date of the Cross of Cong some time in 
the first half of the twelfth century. The cross was 


removed from Tuam to Cong either by Archbishop 
O'Duffy or King Roderic O'Conor, and was found 
there in 1839, when it was purchased by Prof. Mac 
Cullach and presented by him to the Museum of the 
Royal Irish Academy. 

Chalices of earlier date than the Norman Conquest 
are of extreme rarity either in Great Britain or on the 
Continent. Perhaps the three most ancient specimens 
abroad are (i) the chalice, found with gold coins of 
Justinian (a.d. 508 to 527), at Gourdon,^ Chalons-sur- 
Saone, and now in the National Library at Paris ; (2) 
the chalice of Tassilo,"^ Duke of Bavaria (a.d. 757 to 
781), at Kremsmiinster in Lower Austria; and (3) the 
chalice of St. Gozlin^ of Toul (a.d. 922 to 962), now 
in the treasury of the Cathedral of Nancy. The first 
and last of these have two handles. The chalice of 
Tassilo, however, has no handles. It is profusely 
decorated with interlaced-work, zoomorphic designs, 
and figure subjects, and has round the foot the follow- 
ing inscription in capital letters, not unlike those used 
in the Hiberno-Saxon MSS : — 


The lady referred to was Luitberga, wife of Duke 
Tassilo, and daughter of Desiderius, the last king of 
the Lombards. The chalice is 10 inches high and is 
made of copper ornamented with gold, silver, and 
niello. The figures are placed in oval medallions 
round the bowl and the base. The principal figure 
is that of Christ giving the benediction, and the re- 
mainder appear to be those of saints. The style of the 

^ De Caumont's AhecMaire d'Archdologie Architecture Religieuse, 
p. 117. 

'•^ Dr. R. Munro's Boznia-Herzegovina and Dalmatia, p. 292. 
^ De Caumont, loc. cit., p. 118. 


decoration resembles that of the Irish metalwork to a 
certain extent, and the chalice of Tassilo may very 
possibly have been made abroad under the direction 
of some Irish monk. 

Only one metal chalice of undoubted Irish work has 
been preserved until the present time, namely, the 
Ardagh Chalice^ in the Museum of the Royal Irish 
Academy at Dublin. It was found in 1868 in a rath in 
the townland of Reerasta, in the parish of Ardagh, 
Co. Limerick. The chalice belongs to the two-handled 
type, and has a hemispherical bowl, a very short cylin- 
drical stem, and a conical base with a flat rim round 
the bottom. It is 7 inches high by gh inches in 
diameter at the top, and 6h inches in diameter at the 
bottom, the bowl being 4 inches deep and of sufficient 
capacity to hold three pints of liquid. The chalice is 
composed of gold (i oz. 2 dwts.), silver (20 ozs. 
13 dwts.), bronze (9 ozs.), lead, enamel, glass, amber, 
and mica. No less than 354 different pieces, including 
20 rivets, are used in the construction of the vessel. 

The exterior of the bowl of the Ardagh chalice is 
inscribed with the names of the Twelve Apostles in 
Hiberno-Saxon capitals, finely engraved on the silver. 
The forms of the letters correspond with those used in 
the Books of Kells, Dimma, St. Chad, Durham, and 

The raised decoration of the chalice, which is made 
in separate pieces and fixed on with rivets, is concen- 
trated on the following parts : — 

(i) A horizontal band just below the rim and running 
through the handles. 

^ Petrie's Christiari Inscriptions in the Irish Language^ vol. ii., p. 123; 
Trans. R.I.A., vol. xxiv., p. 433; M. Stokes' Early Christian Art 
in Ireland^ p. 83. 









(2) The two handles. 

(3) Two circular medallions on the lower side of the 

bowl midway between the handles. 

(4) The stem. 

(5) The flat rim at the bottom of the base. 

(6) The under side of the flat rim round the base. 

(7) The circular medallion in the centre of the under side 

of the conical base. 

The ornament consists of interlaced-work, step- 
patterns, key-patterns, spiralwork, zoomorphic designs, 
and scrollwork, arranged in panels after the usual 
Celtic fashion. The step patterns are confined to the 
plaques and bosses of enamel, and the other patterns 
are executed in delicate gold filigree-work on a repousse 
background of gold. On the under side of the flat 
rim round the base panels of most beautifully plaited 
silver wire are introduced. Amber is used on the 
handles for the borders round the raised bosses of 
enamel, and there is a narrow ring of the same 
material between the concentric rings of ornament in 
the middle of the under side of the base. The heads 
of the rivets by which the circular medallions on the 
sides of the bowl are fixed are concealed by two small 
bosses of blue glass and two of amber. The heads of 
the rivets for securing the two handles in place are 
disguised in a similar manner. The stem and supports 
of the chalice are of bronze gilt, highly ornamented. 
They are attached to the bowl by a bronze-gilt ball, 
with a strong square tang, and most ingeniously 
fastened by an iron bolt which secures all together. 
A plate of lead is inserted between the upper and 
under sides of the flat rim round the base to give 
weight and stability. The flat rim round the base is 
ornamented with gold and bronze-gilt plaques of open- 


work on a background of mica, in order to show up 
the beauty of the patterns. The flat rim round the 
base has on its under side, between the panels of 
ornament, rectangular tablets of blue glass, under- 
neath which are decorated pieces of wrought-silver, 
which give a brilliant appearance in a strong light. 
In the centre of the under side of the base is a circular 
setting of rock-crystal. The rim of the bowl of the 
chalice is of brass. 

Enough has been said of the elaborate nature of the 
construction and ornamentation of the Ardagh Chalice 
to show that it is a masterpiece of Celtic art metalwork 
of the best period. The style of the lettering of the 
inscription upon it and the general character of the 
decorative features indicate that it belongs to the same 
school as the Book of Kells, the Durham Book, St. 
Chad's Gospels, and the Tara Brooch, and cannot 
consequently be of much later date than the eighth 
century. It will be noticed that in the decoration 
of the Ardagh Chalice spiral patterns of the best 
quality are present, and that the zoomorphs are kept 
under proper restraint so as not to swamp the whole 
design. Both these points are an indication of early 

There are at least three examples known of bronze 
plaques with representations upon them of the Cruci- 
fixion treated in the archaic Irish fashion. The most 
interesting of these was found at Athlone,^ and is now 
in the Museum of the Irish Academy in Dublin. The 
Saviour is shown wearing a tunic, the surface of which 
is almost entirely covered with spirals, key-patterns, 
and interlaced-work. Another smaller and less orna- 
mental plaque with the Crucifixion may be seen in the 

^ Dr. J. Stuart's Sculptured Stones oj Scotland, vol. ii., pi. lo. 


same museum;^ and a third, belonging to Mr. M. J. 
Arketell, has been illustrated by Prof. J. O. Westwood 
in his Miniatures and Ornaments of the Anglo-Saxon 
and Irish MSS. ' 

Leaving Celtic ecclesiastical metalwork, we come to 
personal ornaments, which, although exhibiting the 
same style of decoration, were not necessarily intended 
to be worn by persons taking part in the ceremonies of 
the Church. These personal ornaments consist of 
pins, brooches, and buckles. We have previously 
given a list of the hammer-headed pins, which may 
either be Pagan or Christian. Another peculiarly 
Celtic type of pin consisted of three parts, namel}^, 
(i) a long pin ; (2) a kite-shaped pendant ; and (3) a 
short bar hinged at one end to the top of the pin, and 
at the other to the rounded top of the pendant. A 
remarkably fine pin of this description was found 
about 1883 at Clonmacnois,"^ King's Co., and is now 
in the possession of the Rev. Timothy Lee, of Limerick. 
The pin is 7^ inches long, the coupling-bar \ inch 
long, and the kite-shaped pendant 2J inches long by 
i^ inches wide by \ inch thick. The whole is of silver, 
decorated with gold filigree, enamel, niello, and set- 
tings of claret-coloured glass or precious stone. 
The coupling-bar has on one side a lozenge-shaped 
panel of filigree-work, and on the other an inter- 
laced pattern in niello. The front of the pendant is 
ornamented with a cross having a large rectangular 
setting of glass in the centre, three smaller rectangular 
settings at the ends of the top and two side arms, and 
a small triangular setting at the bottom of the shaft. 
The background of the cross consists of four panels 

^ Miniatures^ pi. 51. Fig-. 7. '^ PI. 51, Fig-. 8. 

'^ Journ. R. Soc. Ant. Ireland^ sen 5, vol. i. (1890-1), p. 318. 


of interlaced filigree-work, three of which are missing. 
The point of the kite-shaped pendant terminates in a 
beast's head. On the back of the pendant there is 
a cross of similar shape to that on the front, but with 
an ornamental border of spiralwork round it, and the 
whole design executed in niello. At the pointed end 
at the bottom is fixed a small ring through which 
passes a silver plaited chain of Trichinopoly-work, 
like the one attached to the Tara Brooch. There is 
another pin of similar shape ornamented with zoo- 
morphic designs in the Museum of the Royal Irish 
Academy,^ in Dublin. 

Dr. Hans Hildebrand, in his excellent South Ken- 
sington handbook of The Industrial Arts of Scandmavia 
(p..2i), remarks that "every work of human art, higher 
as well as lower, has its shape determined by two 
agents : the end which it is to serve, and the taste of 
the people and the time of which it is a fruit." In 
other words, there is a utilitarian as well as an orna- 
mental side to almost every object fashioned by man to 
satisfy his wants. The form of an object must depend 
primarily upon the practical use to which it is intended 
to be put, and the decorative features generally follow 
afterwards in due course. The function of the deco- 
rative features, however, should be to add grace and 
beauty to the original form of the object, but not to 
attempt to disguise the utilitarian purpose it fulfils. 

No relics of antiquity are more deserving of study 
than personal ornaments, and of all personal orna- 
ments perhaps the brooch is the most important as 
affording an insight into the character of the people by 
whom it was worn. Their ingenuity can be measured 
by the perfection of the mechanism of the working 

^ R.I. A. photo, A {65. 

Pin-brooch from Clonmacnois, King-'s Co. 

Now in the possession of the Rev. Timothy Lee, of Limerick 
Drawn by R. Cochrane, f.s.a. _, 


parts, their culture by the refinement of the ornament, 
and their skill as craftsmen by the finish of the work- 
manship. Much, again, is to be learnt of the habits of 
the people by investigating the different methods of 
wearing the brooch. Thus it is that almost every age 
and every country possesses its typical form of brooch. 

Looked at from its practical side, a brooch is a con- 
trivance for fastening together temporarily any two 
points on a garment. It is obviously a higher de- 
velopment of the pin. Going back to first principles, 
the pin may have been suggested by the natural spikes, 
or thorns, found in the vegetable world. It would not 
require much intelligence to see that a small knob 
added to the blunt end of the pin would facilitate its 
removal from the fabric when it was required to be 
withdrawn, and would also prevent the pin going 
further than was desirable through the fabric. The 
problem which was solved by the invention of the 
brooch, however, was one of much greater complexity, 
namely, how to secure the pin in position so as to 
prevent it from slipping out of the fabric in the direc- 
tion of the head. This might have been effected either 
by fixing a removable knob, or stop of some kind, on 
the pointed end after it had been inserted in the fabric, 
or by connecting the head with the point temporarily, 
so as to form a complete ring for the time being. In 
the brooch the latter alternative is chosen. The pin 
must necessarily be straight, so as to pierce the fabric 
with the least amount of resistance, and the temporary 
connection between the head and the point has to be 
approximately semicircular, the whole forming a ring 
shaped like a bow, the pin corresponding to the string 
and the body of the brooch to the bow. 

In order to be able to remove the brooch from the 



3 3, 3 J 5 ,^ 3 

1 3 -; , 3 3 3 ,3-1 

3 J 3 

■'3 ■' 3 

3 3 2 


fabric at pleasure, some contrivance must be hit upon 
by which a gap, or break, can be made in the ring, and 
be closed up again whenever it is desired to do so. 
The opening is attained by placing a hinge where the 
head of a pin joins the body of the brooch, and the 
closing by having a groove-shaped catch at the opposite 
extremity. A spring is also required to prevent the 
pin coming unfastened accidentally from the catch. 
These different contrivances constitute the essential 
parts of a brooch, which, divested of its ornamental 
appendages, is represented by the ordinary "safety- 
pin " of the present day. 

If the rigid bow-like connection between the head 
and point of the pin be doubled we get an annular 
brooch, and if the central portion of the ring be filled 
in we get the discoidal brooch. In these cases the ring 
or disc is placed parallel to the plane of the fabric 
instead of at right angles to it. 

The somewhat dry disquisition just inflicted upon 
the unsuspecting reader is necessary in order to place 
him in a position to fully understand the mechanism of 
the typical Celtic brooch, the leading characteristics of 
which are that the ring has a break in its continuity 
(whence the name " penannular "), and that the length 
of the pin considerably exceeds the diameter of the 
ring. The object of the break in the continuity of the 
ring is that it enables the spring-catch to be dispensed 
with, the method of fixing the brooch in the dress 
being as follows : First, the long pin is inserted in 
the fabric at two points close together, in such a 
manner that the apex goes right through it and appears 
again above the surface ; the pin is then forced through 
the break, and the ring is given a turn through a right 
angle in the plane of the fabric, thus fixing the brooch 


by the friction produced by the drag of the weight of 
the garment on the pin. 

We are now brought face to face with the question 
as to how the Celtic penannular brooch was worn. 
This can not only be conjecturally determined by an 
examination of the specimens to be found in museums, 
but fortunately can be settled beyond a shadow of a 
doubt in two ways, each of which confirms the other. 
First, there are at least two contemporary representa- 
tions of persons actually wearing a penannular brooch 
(one on a cross at Monasterboice, Co. Louth, and the 
other on a cross at Kells, Co. Meath, in Ireland) ; and 
this ancient form of fibula has survived, and is in use 
at the present time in Algeria and elsewhere. 

The example at Monasterboice^ is on the bottom 
panel of the side of the shaft of the cross of Muiredach 
(or Murdoch), which was erected in a.d. 924. The 
scene represented on the panel has been conjectured by 
the late Prof. J. O. Westwood, from its similarity to a 
miniature in the Book of Kells at Trinity College, 
Dublin, to be intended for Christ seized by the Jews. 
If this be so, the central figure is our Lord, and on 
each side is a soldier armed with a drawn sword. The 
sculpture is in good preservation, considering its great 
age, and the details of the costume, which are very 
elaborate, can be made out fairly well. Our Lord 
wears a sort of cloak with a penannular brooch fixed 
on His right shoulder. The split in the ring of the 
brooch faces downwards, and the pin is inclined up- 
wards at an angle of about 30 degrees to the horizontal, 
the point being outwards. Probably the heavy head 
of the pin is placed downwards because its weight 
would always tend to bring it to this position, as the 

^ Illustrated Anhceoloirist *io\: 1893, p. 164. 




one of most stable equilibrium, but it may also have 
been to avoid injury from the point of the long 

The second example is on the bottom panel of the 
side of the broken cross-shaft in Kells^ churchyard. 
The exact date of this monument is unknown, but it 
is probably of the ninth or tenth century. The subject 
on the panel is the Baptism of Christ, with the sources 
of the two imaginary rivers, Jor and Dan, which, when 
united, were supposed to contribute their waters to the 
Jordan, indicated conventionally in a most remarkable 
manner. John the Baptist pours the water over the 
head of Christ with a sort of ladle. Above is the Holy 
Dove, and on the left are two figures wearing pen- 
annular brooches exactly in the same manner as on the 
Monasterboice cross, with the pin pointing upwards. 
In the case of the figure furthest to the left, the end of 
the long pin is inserted a second time into the fabric 
of the dress, beyond the ring. 

The method of wearing the penannular brooch at the 
present day in Algeria is clearly indicated on the re- 
production of a photograph" here given. The only 
difi'erence in the way of wearing the brooch in Algeria 
and in ancient Ireland is, that in the former case they 
are worn in pairs instead of singly, and there is a 
connecting chain with a small pendant scent-box hung 
from the middle. The size of the box is exaggerated 
out of all proportion by being placed nearer the camera 
than the rest of the figure. 

In Great Britain the penannular brooches appear to 
have been worn singly, as they are never found in 

^ Illustrated Archceologist ^ov 1893, p. 165. 

^ Obtained from Albert Hautecoeur, 2, Boulevard des Capucines, 


pairs ; thus offering a contrast to the Scandinavian 
bowl-shaped brooches, which are always found in pairs, 
and were connected by a chain, as in the case of the 
Algerian brooches. 

It would be interesting to know how the penannular 
form of brooch was first introduced into this country, 
for its seems hardly conceivable that it could have 
been invented here, or else it would not be found in 
Algeria, which never had any connection with Great 
Britain, it being extremely unlikely that so peculiar a 
type of brooch was evolved independently in the two 

The most probable suggestion is that the Algerians 
and the ancient Irish got it from a common source, 
namely, the East, and that its introduction into our 
own islands dates from the time when the traffic in 
silver bullion from the East commenced. The existence 
of a trade route which was made use of by the dealers 
in silver bullion is made clear by the number of finds 
of Mahomedan silver coins associated with ingots, 
rings, and ornaments of silver, made both in Scandina- 
via and in Great Britain. Dr. Hans Hildebrand, in his 
Industrial Arts of Scandinavia (p. 8i), informs us that 
'^considerable stores of such coins, most of them of 
the Samanid dynasty, have been found in Sweden. It 
is satisfactorily proved by Russian finds, that these 
coins were brought from states near the Caspian Sea, 
through Russia to the shores of the Baltic Sea, and 
thence to the commerce established by the inhabitants 
of Gotland over to that island. From Gotland, and 
probably also by direct intercourse with Russia, the 
Mahomedan coins were spread over Scandinavia, being 
of course more common in the eastern provinces of 
Sweden than in the western and in Norway." No 




less than 20,000 Mahomedan silver coins have already 
been discovered in Sweden, mostly dating between 
A.D. 880 and 955, the latest belonging to the year 

A.D. lOIO. 

Penannular brooches have been found in association 
with Mahomedan coins of the ninth and tenth cen- 
turies, at Skaill, in Orkney ; at Storr, in Skye ; and at 
Cuerdale, near Preston, in Lancashire. 

Although the general form of the penannular brooch 
is probably of Eastern origin, the decorative features 
vary according to the race of people who adopted it. 
Thus the examples from Algeria have Mahomedan 
ornament; those from Gotland, Scandinavian patterns; 
whilst those from Ireland and Scotland are thoroughly 
Celtic in design. With the decoration of the foreign 
specimens we are not now concerned, but a few words 
with regard to the various types found in Great Britain 
will form a fitting conclusion to this article. 

The finest collections of penannular brooches are to 
be seen in the British Museum, the National Museum 
of Antiquities of Scotland, in Edinburgh, and in the 
Museum of the Royal Irish Academy in Dublin. A 
few good specimens are in private hands, and there is 
a splendid one from Orton Scar,^ in Westmoreland, in 
the Museum of the Society of Antiquaries at Burling- 
ton House. 

The portions of the brooch, the forms of which are 
altered so as to adapt them better to the reception of 
ornament, are the head of the pin and the two termina- 
tions of the ring, where the break occurs. The two 
chief ways of altering the shapes of these parts are 
(i) by making them spherical, and (2) by expanding 
into a wide flat surface ; the object in both cases being 

^ Reliquary for 1903, p. 203. 


to increase the area available for decoration. Some- 
times, also, the ring and the long end of the pin are 
flattened and widened for a similar purpose. 

As an example of a penannular brooch with bulbous 
terminations to the ring and head of the pin, we have 
one from Co. Kildare in Ireland (R.I.A. photo, B 172). 
The knobs are covered with a prickly ornament pro- 
duced by incised lines drawn diagonally in two 
directions, crossing each other, giving the whole 
the appearance of the head of a thistle. Several 
brooches of this kind have been obtained from difterent 
localities in Ireland, and there was one along with the 
three brooches of the type with flattened and expanded 
ends found with the Ardagh Chalice — a hoard of 
objects of purely Irish types — but their ornamentation 
appears to be more Scandinavian than Celtic. One 
of the best specimens from Skaill, in Orkney, now in 
the Edinburgh Museum, has a pin i foot 3 inches 
long, and the bulbous ends covered with zoomorphic 
designs similar to those on the Manx crosses, and on 
an iron axe-head inlaid with silver from the Mammen 
How,i Denmark. 

We next come to brooches with discoidal termina- 
tions, of a date not later than the beginning of the 
ninth century, as the simplest example of which may 
be taken one from Croy, in Inverness-shire {Scotland 
in Early Christian lYmes^ 2nd ser., p. 23). Another, 
found near Perth {ibicL^ p. 21), has three raised heads 
on each disc ; whilst one from Rogart, in Sutherland- 
shire {ibid,^ p. 7), has four raised heads outside the 
circumference of the disc, so that the terminations are 
altered into the shape of a quartrefoil. 

Lastly, we have brooches with flat expanded ends to 

1 Dr, J. Anderson's Scotland hi Pagan Times : Iron Age, p. 97. 





the ring, of which kind three specimens in the Museum 
of the Royal Irish Academy at Dublin are illustrated, 
in order to show the way of ornamenting the expan- 
sions with one, four, and five raised bosses, having 
zoomorphic designs on the background (R.I. A. photos, 
B 163 and B 164). The area of the head of the pin 
available for decoration is increased by making it into 
a cylindrical tube. 

In the final stage of the development of the pen- 
annular brooch in Ireland it ceased to be penannular, 
if we may be permitted to use such an Irish expression. 
The break in the ring was entirely filled up, although 
its position can still be traced by the method of 
arranging the pattern, which survived in its old form 
long after the split had disappeared. The celebrated 
Tara Brooch, in the Museum of the Royal Irish 
Academy (R.I.A. photo, A 161), affords a striking 
example of this. The doing away of the break in the 
ring must have entirely defeated the original purpose 
the brooch was intended to serve, and it would, there- 
fore, appear that these highly decorated brooches were 
made rather for ceremonial use, than to be of any 
practical value as dress-fasteners. 

It may be pointed out that all the characteristic 
modifications of the form of the penannular brooch 
made by the Celtic artist arose from his desire to 
provide more space for the ornamental patterns, which 
were the very salt of his existence. 

Dr. Joseph Anderson contributes the following note 
apropos of the long pin : — 

" In the Brehon Laws, vol. iii., p. 291, men are exempted 
from liability to fine for injury from the pin of their brooch 
(in a crush? or at a fair?) if they have the brooch on their 
shoulder so as not to project beyond it. Women also are 


exempt if they have their brooch similarly on their bosom." 
Vol. iv. , p. 323, "a precious brooch worth an ounce [of 
silver?] is enumerated among the customary insignia of 
a chief." 

The Tara Brooch^ was found in 1850 by some 
children w^hilst playing on the strand near Drogheda, 
Co. Meath. It was offered by the mother of the 
children to a dealer in metals in Drogheda, but he 
refused to purchase it, after which she took it to a 
watchmaker in the town, who gave her a trifle for it. 
The watchmaker cleaned it up, and subsequently sold it 
to Messrs. Waterhouse, of Dame Street, Dublin. The 
Tara Brooch is now in the Museum of the Royal Irish 
Academy. The body of the brooch is made of an 
alloy of copper and tin called white bronze, and the 
decorations with which it is encrusted consist of gold 
filigree in small recessed panels, niello, enamel, and 
settings of amber and glass. The ornament includes 
interlaced-work, spirals, step-patterns, scrollwork, zoo- 
morphs, and anthropomorphs. The spiralwork is of 
the best kind, such as is only found in MSS. like 
the Book of Kells. The designs on the back of the 
brooch appear to be chased or cut into the solid metal 
of the body, and not composed of plaques fixed on 
with rivets. Attention should be particularly directed 
to the rows of birds, each biting the leg of the one in 
front of it, on the back of the brooch. Similar designs 
occur in the Lindisfarne Gospels' and on a cross-shaft 
from Aberlady,^ now at Carlowrie Castle, near Kirk- 

^ H. O'Neill's Fhie Arts of Ancient Ireland^ p. 49. 

2 Publication of the Palceographical Soc, and G. F, Warner's Ilhi- 
viinated Manuscripts in the British Mnseuni, 3rd series. 

■^ Allen and Anderson's Early Christian Monuments of Scotland^ 
p. 42S. 

^ O 


o o 


listen, Linlithgowshire, clearly showing Northumbrian 
influence, as bird-motived ornament of this kind is in 
no way characteristic of pure Irish work. 

There are several beautiful penannular brooches in 
the National Museum of Antiquities of Scotland, at 
Edinburgh, most of which are described and illustrated 
in Dr. J. Anderson's Scotland in Early Christian Times 
series. The finest of these is the Hunterston Brooch,^ 
which has a Runic inscription upon it and is deco- 
rated with interlaced-work, zoomorphs, and spiralwork 
almost equal to that on the Tara Brooch. The Cadboll 
Brooch^ from Rogart, Sutherlandshire, and a brooch 
from Perth'^ are also very beautiful examples. 

The best examples of early Irish ornamental leather- 
work are the satchel of the Book of Armagh^ in the 
Library of Trinity College, Dublin, and the satchel of 
St. Moedog's reliquary"^ in the Museum of the Royal 
Irish Academy. The patterns on the former consist of 
interlaced-work and zoomorphs, and those on the latter 
of interlaced-work only. There are also specimens of 
leather shoes in the Museum of the Royal Irish 
Academy with Celtic ornament upon them.'^ 

There are very few objects of wood or bone now in 
existence which exhibit Celtic ornament of the Christian 

^ Dr. J. Anderson's Scotland in Early Christian Times, 2nd ser. , p. 2. 

'■^ Ibid., p. 7. 

^ Ibid., p. 21. 

^ Rev. J. P. Mahaffy's Book of Trinity College. 

^ Archcelogia, vol. xliii., p. 13 r. 

6 Sir W. Wilde's Catal. Mus. R.I. A., p. 284. 



ECCLESIASTICAL and other MSS. written on 
sheets of vellum and bound up in the form of 
a book were introduced into this country with 
Christianity. The materials and tools used by the 
Celtic scribes and illuminators probably did not differ 
to any great extent from those used throughout 
Europe during the Middle Ages. The parchment of 
the Irish MSvS. is, however, generally much thicker ^ 
than that of the Carlovingian and other foreign MSS. 
The letters in the Irish MSS. of the best period, such 
as the Book of Kells, are composed partly of extremely 
fine lines, drawn with a firm hand, which gradually 
expand in width to form the other parts of the letters. 
These could hardly have been made with a reed or a 
brush, so that it is probable that the pens of the Irish 
scribes were made from the quills of swans, geese, 
crows, and other birds. The black ink used in the 
Irish MSS. is remarkable for its blackness and dura- 
bility ; and Bede, the historian, speaks highly in praise 
of the colours prepared in Ireland, and especially of 

^ Sir E. M. Thompson's Greek and Lalui Palcpograp/iy, p. 38. 



the brilliancy and permanence of the red made from 
whelks.^ Some colours, such as yellow, are put on 
thin and transparent ; whilst others, such as red, have 
a thick body made of titurated earth or other skilfully 
prepared material, mixed with some strong binding 
material of the nature of gum or varnish.^ 

The material employed for the highly ornamented 
sculptured monuments of the Christian Celtic period 
was generally that most readily procurable on the spot, 
but a preference was always shown for a freestone, 
which could be easily worked. The greater proportion 
of the best crosses are carved in a fine-grained sand- 
stone. In Cornwall granite was most generally used, 
although Polyphant stone was also used. In the Isle 
of Man nearly all the crosses are of slate. Hard, 
volcanic rocks were avoided where possible on account 
of the difficulty of working. There are, however, 
crosses of trap-rock at Carew, Pembrokeshire, and 
Moel Siarman, Brecknockshire. 

On some of the crosses the marks of the tool with 
which they were carved can still be clearly seen, as on 
the Cross of Iltyd at Llantwit Major, and the cross- 
base at Llangevelach, both in Glamorganshire. As 
far as it is possible to judge from the tool-marks, either 
a pick or a pointed chisel must have been employed 
by the early Christian Celtic stone-carvers. Similar 
tool-marks have been observed on the cup-and-ring 
sculptures of the Bronze Age. 

In the churchyard at Kells, Co. Meath, there is an 
unfinished cross which is of great interest as showing 
the exact methods used in the construction and decora- 
tion of this class of monument. The stone was first 

^ Bede's Ecch Hist., bk. i,, chap, i, 

'■^ Dr. Ferdinand Keller in the Ulster Journal of ArchcEology, vol. viii. 


squared and the design roughly set out upon it. 
Draughts were then cut across the faces, leaving 
certain portions standing out in high relief, upon 
which the figure subjects were afterwards sculptured. 
The unfinished cross at Kells was formerly lying on 
the ground, but it has recently been erected on its 
original base, which is also unfinished. 

When the crosses are constructed of two or more 
pieces they are fitted together by means of mortice 
and tenon joints. Sometimes the quadrants of the 
circular ring connecting the arms were made in sepa- 
rate pieces, as in the case of the large broken cross 
at lona.^ 

The metals in use during the Christian Celtic period 
were gold, silver, copper, lead, bronze, brass, and 
other alloys. These were cast and wrought and 
ornamented by means of enamelling, niello, plating, 
gilding, repousse-work, chasing, engraving, piercing, 
inlaying, filigree-work, Trichinopoly chainwork, and 
settings of precious stones, amber, and glass. The 
different pieces of the metal objects were fixed together 
by rivets, and if soldering and brazing were known, 
they were certainly not employed to any great extent. 
Even when the specimens can be removed from their 
show-cases in museums and examined carefully by an 
expert it is not always possible to be certain of the 
exact technical processes by which the various decora- 
tive effects have been produced, and unless the objects 
can be dissected many of the constructive features must 
necessarily be a matter for conjecture. The Ardagh 
Chalice and the Tara Brooch illustrate nearly all the 
materials, technical processes, and methods of con- 
struction used at this period. 

^ Proc. Sac. ^lut. Scot., vol. xxxv. , p. 90. 


Three different kinds of enamel are used in the 
decoration of the Ardagh Chalice, namely, (i) a 
peculiar variety of cloisonne in which the compartments, 
or cloisons, are all cut out of a single piece of metal 
and the open framework thus formed is pressed into 
the surface of the enamel when soft until it rises up 
and fills each compartment ; (2) a combination of 
cloisonne and champleve enamel in which the com- 
partments are all cut out of a single piece of metal, 
some being pierced right through and the remainder 
only sunk partially through the thickness of the metal; 
the framework is pressed into the enamel when soft, 
thus filling up the open compartments, as in the first 
kind just described, and the remaining dug-out com- 
partments are filled with fusible enamel as in champleve; 
and (3) a species of champleve enamel, in which the 
surface of a piece of glass was engraved with a design 
in intaglio and the hollows filled up with an enamel of 
a different colour. The Celtic enamels of the Christian 
period usually occur in the form of small round bosses, 
of which there are good instances on the Ardagh 
Chalice, the Ardagh Brooch, the Tara Brooch, the 
Lismore Crozier, and the Cross of Cong. 

The use of bands of silver with borders of niello is 
well illustrated by the head of a crozier^ formerly 
belonging to the late Dr. W. Frazer, m.r.i.a., of 
Dublin. Portions of the silver have been stripped, 
showing how the surface of the metal into which 
it was inlaid was roughened with a pointed tool to 
make the inlay adhere better. Niello is a black com- 
position made of silver, lead, sulphur, and copper, 
which is reduced to powder and placed in cavities or 
lines cut for its reception in the surface of the metal, 

^ Proc. R.I. A., 3rd sen, vol. i., p. 207. 


and afterwards incorporated with it by being passed 
through the furnace. Niello probably found its 
way to Ireland from the East. It was used by the 
Byzantines as early as the beginning of the ninth 
century. 1 

A peculiar kind of decoration which is specially 
characteristic of the early Irish ecclesiastical metal- 
work consists of plates perforated with triangles, 
squares, and crosses, so as to form a geometrical 
pattern. The plates are usually of bronze covered 
with silver, and the contrast between the bright surface 
of the white silver and the pierced portions through 
which the dark bronze below can be seen gives the 
general appearance of chequerwork. There are good 
instances of this class of decoration on the Shrine of 
the Bell of St. Patrick's Will, and the Cumdachs 
of Dimma's Book, the Stowe Missal, and the Shrine 
of St. Mogue. Cruciform pierced work of a similar 
kind also occurs on an ivory of the tenth century 
representing the raising of the widow of Nain's son, 
in the British Museum ;'' on an ivory of the tenth 
century, representing Christ in the Temple, in the 
Royal Library at Berlin ;^ and on the chair of the 
image of St. Faith, in the treasury of Conques^ 
(Aveyron). The wards of ecclesiastical keys are often 
made to form cruciform patterns, as in the case of 
those of Netley Abbey, St. Serrais Maastricht, and 
Liege. -^ The cruciform patterns on the west face of 

^ J. H. Pollen's Go/d and Silver, p. 53. 

- J. O. Westwood's Catal. of Fictile Ivories in S. K. Mus. 

■■•' Ibid. 

■^ Annales de la Saddle Archdologiqiie de Briixelles, vol. xv, (1901), 
P- 434- 

° Le Chanoine Rensens' Eldnents d' Archdologie Chrdtienne, 2nd ed. 
(Aix, 1885), vol. i., pp. 241 and 262. 


the cross at Dysert O'Dea,^ Co. Clare, seem to be 
copied from metalwork. 

Filigree-work of gold wire is used to make the 
panels of interlaced-work, scrollwork, and zoomorphic 
designs with which some of the best specimens of 
Christian Celtic metalwork are decorated, such as the 
Ardagh Chalice, the Tara Brooch, the Hunterston 
Brooch, and the Clonmacnois Pin. The filigree-work 
is often covered with minute granulations, which add 
greatly to the richness of the effect produced by their 

We have already referred to the Trichinopoly chain- 
work of silver wire used in the Ardagh Chalice, the 
Tara Brooch, and the Clonmacnois Pin. This kind of 
chainwork can be traced back to the Pagan Celtic 
period, as chains of similar character were found with 
the Late-Celtic gold collar at Limavady, Co. London- 
derry, and with the pair of silver-gilt Kelto-Roman 
fibulae from Chorley, Lancashire, now in the British 

Settings of coral and enamel were, as we have seen, 
employed for the decoration of the Late-Celtic metal- 
work, but in the Christian Celtic period numerous 
other substances were also employed, such as glass, 
rock crystal, amber, and other precious stones. In 
some cases the settings of stones and glass were rect- 
angular with a flat top and bevelled edges, but they 
were more generally round, oval, or almond-shaped 
and ^^ tallow-cut," i.e. polished without facets. 

The process used for producing the patterns on the 
leather satchels and shoes previously mentioned was 
probably of the same nature as that by which the cuir 
houilli cases of later times were decorated. 

^ Jour. R, Soc. Ant, Ireland, ser. 5, vol. ix., p. 251. 


Objects of wood, bone, ivory, and pottery and textile 
fabrics of the Christian Celtic period are so rare that 
there is really nothing to be said about the technical 
processes involved in their manufacture. 


Attention has been recently directed to the problem 
of how decorative art was evolved, in the first instance, 
by the primitive races of mankind in remote ages. 
Mr. Henry Balfour, Mr. C. H. Read, and Dr. Colley 
March have shown us how much light may be thrown 
on this difficult question by a critical examination of 
the various forms of ornament used by the savage — or, 
rather, the uncultured — peoples existing at the present 
day in countries where they have had only limited 
opportunities of coming in contact with modern civili- 

There is, however, at least as difficult a problem 
nearer our own doors awaiting solution, namely, that 
of the origin and development of early Christian 
decorative art in the British Isles. This problem is 
not one of a wholly uncultured race left to itself to 
work out its own ideas, as suggested by external 
natural objects or otherwise, but it is a problem of a 
race already in a state of semi-culture being brought 
suddenly face to face with a higher civilisation, through 
the introduction of a new religion, and afterwards 
influencing, or being influenced by, other conquering 
races — also in a state of semi-culture — whom they con- 
verted by missionary enterprise. That is to say, the 
Celts of Ireland, Scotland, the Isle of Man, Wales, 


and Cornwall became acquainted with Italo-Byzantine 
art when they were first Christianised, about the middle 
of the fifth century. In the seventh century they came 
in contact with the Anglo-Saxons, and in the ninth 
with the Norsemen and Danes. It is the object of 
the present inquiry to determine in what measure the 
Christian art of this country before the Norman Con- 
quest was affected by the absorption of these new 
racial elements. 

The style of art we are now dealing with was 
formerly, quite wrongly, called Runic, because some 
of the monuments on which characteristic forms of 
ornament occur bear Runic inscriptions. Later au- 
thorities have called the style Hiberno-Saxon, Kelto- 
Northumbrian, Celtic, and Irish, but this is simply 
begging the whole question. The term we have 
chosen, namely, early Christian, is scientifically cor- 
rect, and does not commit us to the assumption of any 
unproved facts. 

Early Christian art in this country is essentially 
decorative, and to a lesser extent symbolic. The 
figure subjects are obviously barbarous copies of 
Byzantine originals, for no matter how they are dis- 
guised by bad drawing or incrusted with ornament, 
the conventional grouping and accessories still remain 
to prove their origin. The miniature of the Temptation 
of Christ in the Book of Kells is perhaps the most 
remarkable instance of a Byzantine design Celticised, 
if one may use the expression. Comparing this with a 
miniature in the Psalter of Misselinda (a.d. 1066) in 
the British Museum (Add. 19,352), we find all the 
essential features of the scene the same, even to the 
black Devil ; but in the Book of Kells the Temple 
with its Byzantine cupolas has been converted into an 


Irish stone-roofed oratory, shaped like a metal shrine 
of the period, and covered with ornament ; the Devil, 
too, has been decorated with spiral curves after the 
Celtic fashion. 

The miniatures of the Evangelists, with their sym- 
bols, which form the frontispieces of the Irish Gospels, 
are also taken from a Byzantine source and similarly 
disguised, although not so effectually as to conceal 
their derivation. The Irish illuminator put as much 
local colour into his copy as a Chinaman or a Japanese 
would, but in a different way, if told to make a replica 
of an English picture. 

In distilling the original Byzantine idea through the 
alembic of the mind of the Irish scribe it has absorbed 
so much of his individuality that it assumes an archaic 
and semi-barbarous appearance which is very mis- 
leading at first sight. We hope to be able to show 
that some of the elements of the ornament may be 
traced to a Byzantine source, and that the only obstacle 
in the way of our at once recognising whence the Irish 
designer received his inspiration is his marvellous 
power of adaptation and skill in evolving fresh com- 
binations of simple elements. The ancient Irish artists 
appear in some respects to have resembled the Japanese 
in the rapidity with which they absorbed new ideas 
and turned them to good account in their decorative 

The materials available for the study of early 
Christian art in Britain consist of illuminated MSS., 
ecclesiastical and other metalwork, sculptured monu- 
ments, and a few miscellaneous objects. I propose 
now to direct attention chiefly to the sculptured 
monuments, because they afford a much more certain 
means than any other of determining the charac- 


teristics of the various local styles throughout the 

If a monument is found in a particular district, it may 
generally be assumed that it was the art product of the 
district, unless there is some special reason for thinking 
otherwise. The number of MSS. and examples of 
metalwork is comparatively much smaller than the 
number of monuments, and it is only in a few ex- 
ceptional cases that a MS. can be traced to the 
monastic establishment where it was written. In 
Scotland, for example, although richer than any other 
part of Great Britain in sculptured monuments, the 
Book of Deer is the only pre-Norman MS. known to 
have been written there. Wales, again, can only 
claim the Psalter of Ricemarchus. 

I am of opinion that if we are ever to arrive at any 
definite conclusions with regard to the evolution of 
early Christian art in Great Britain, it must be by 
means of a careful examination and comparison of the 
minute details of the ornament. The science of palaeo- 
graphy is entirely founded on the observation of every 
small variation in the form of each letter, and if the 
same trouble was taken with ornament equally valuable 
results would be obtained. 

We will now proceed to analyse the decorative 
features of the monuments, and endeavour to find 
an origin for the component elements which go to 
make up the style. I must assume the reader to 
possess a certain amount of acquaintance with the 
art of the early Christian period, and to know what 
is meant by most of the technical terms, but I shall 
give examples of the various classes of patterns in case 
anyone should be unfamiliar with their appearance. 

Broadly speaking, early Christian ornament in Great 



Britain is made up of the following elements, generally 
arranged in separate panels : — 

(i) Interlaced-work. 

(2) Step-patterns. 

(3) Key-patterns. 

(4) Spirals. 

(5) Zoomorphic Designs. 1 Suggested by Animal, 

(6) Anthropomorphic Designs, r Human, and Vege- 

(7) Phyllomorphic Designs. J table Forms. 

Now the question is, what are the possible or prob- 
able sources whence each of these different kinds of 
patterns was derived? 

First of all, there are the native and imported styles 
of decorative art existing in Great Britain previous to 
the introduction of Christianity (circa a.d. 450), com- 
prising the art of the ages of Stone, Bronze, and Iron, 
and Romano-British art. Next, the external influ- 
ences which came into play after a.d. 450, and before 
A.D. 1066, were Italo-Byzantine, Anglo-Saxon, Prankish, 
and Scandinavian. 

Early Christian art in Great Britain was produced, 
in the first instance, by grafting the Italo-Byzantine 
style upon the native style of the Iron Age (sometimes 
called Late-Celtic), and was subsequently modified by 
Anglo-Saxon and Scandinavian influence. 

Of the forms of decoration used in the Stone Age in 
this country we know hardly anything, and therefore 
they will not come within the scope of our investiga- 
tions. The ornamental patterns of the Bronze Age, 
as far as we are acquainted with them from a study of 
the sepulchral urns, implements, personal ornaments, 
and sculptured cists and chambered tumuli, are of a 
very simple description, consisting chiefly of chevrons, 
concentric circles, and rudely drawn spirals. The latter 


may have been the forerunners of the beautifully de- 
signed volutes of the Iron Age, the nearest approach 
to perfection being on the sculptured slab at the 
entrance to the New Grange tumulus, Co. Meath, 
and on the slabs forming the sides of a chambered 
cairn at Clover Hill, Co. Sligo. 

When we come down to the Iron Age we find a very 
beautiful and refined system of decoration applied to 
bronze objects, such as hand-mirrors, shields, helmets, 
sword-sheaths, and horse-trappings, the leading motif 
of which is the divergent, or trumpet-shaped spiral. 
This style of decoration has received the name Late- 
Celtic in this country, and La Tene on the Continent. 

No one can fail to be struck with the similarity 
between the Late-Celtic spiral ornament and that found 
in the early Irish MSS., the patterns in some cases 
being absolutely identical. It is thus possible to trace 
this particular element in the decorative art of the early 
Christian period to a native Pagan source. 

Late-Celtic objects have been found in all parts of 
the United Kingdom, but probably the style of decora- 
tion only survived into Christian times in Ireland, 
although there is really no reason why it should not 
have done so elsewhere — in the north of Scotland, 
for instance, which was quite as much cut off from 
civilisation as Ireland during the Saxon conquests. 
The closest resemblance between the spiral decoration 
of the Pagan period and that of the Christian period 
is to be found on the discoidal ornaments with pat- 
terns in champleve enamel, forming the attachments of 
the handles of certain bronze bowls, several examples 
of which have been discovered from time to time in 
different parts of England.^ 

^ ArchcEologia^ vol. Ivi. , p. 43. 



I believe that the only element in early Christian 
decorative art in this country that can be traced to a 
native Pagan source is the divergent spiral. It has 
been suggested that the Irish and Saxon designers 
derived some of their ideas from the Roman pave- 
ments, but I can see nothing in the decoration of the 
MSS. on monuments of the pre-Norman period that 
can be fairly attributed to a Romano-British origin. 

We have now to consider the external influences 
which came into play after the introduction of Chris- 
tianity {circa A.D. 450). First amongst these was the 
influence of Italy, and thus more indirectly that of 
Byzantium. It is to this source that it is possible to 
trace the interlaced-work and scrolls of foliage which 
occur so frequently on the early sculptured monuments 
in Great Britain. We can refer to no better text-book 
whilst dealing with this portion of our investigation 
than L^ Architettura in Italia, by Professor Rafl'aele 
Cattaneo (Venezia, 1888), who, by a careful study of 
the subject, has been able to divide early Italian 
ecclesiastical architecture into the following styles and 
corresponding periods : — 

(i) Latino-Barbaro . . a.d. 300 to a.d. 600. 

(2) Bizantino-Barbaro . .a.d. 600 to a.d. 800. 

(3) Italo-Bizantino . .a.d. 800 to a.d. 1000. 

As an example of the first period we have the 
Ciborium in the Church of San Clemente at Rome 
(a.d. 514-23J, decorated with plaitwork and foliage, 
both evidently of Classical origin. Belonging to the 
second period we have the Ciborium of San Giorgio 
di Valpolicella^ (a.d. 712), decorated with broken plait- 

^ Also the jambs of the doorway of the chapel of S. Zeno in the 
church of S. Prassede, Rome (a.d. 772-95). 




work, and the Baptistery of Cividale (a.d. 737), with 
fully developed knotwork. And belonging to the third 
period the Ciborium of Sant' Apollinare in Classe, 
near Ravenna (a.d. 806-16), with interlaced-work and 

Pierced Marble Screen at Raveni 

foliage, and a slab over the altar of San Giacomo, at 
Venice (a.d. 829), with circular knotwork.^ 

A careful examination of these specimens shows that 

^ Slabs of circular knotwork are also to be seen in the church of 
Sta. Sabina, Rome. 


the plait was the first kind of interlaced-work employed 
for decorative purposes, and that it was of Classical 
origin. The plait as a decorative motive must have 
been well known to the inhabitants of this country 
during the Roman occupation and immediately after, 
from the numerous examples which occur on Roman 
pavements, as at Lydney Park, Gloucestershire, and 

Knotwork was gradually evolved from the plait 
by introducing breaks at regular intervals during 
the Bizantino-Barbaro period (a.d. 600 to 800) ; and 
subsequent to this we find still more complicated forms 
of interlaced patterns were introduced, which I propose 
to call circular knotwork and triangular knotwork. The 
evidence gathered from dated examples of interlaced- 
work in Italy tends to show that there was a gradual 
advance in the elaboration of the patterns as time 
went on. Consequently the style could not have 
been borrowed en bloc by Ireland from Italy, or vice 
versa, at one time ; but interlaced ornament must have 
been a prevalent form of decoration throughout the 
whole of the West of Europe, and the style advanced 
in all the different countries simultaneously, there being 
always a constant communication between Rome and 
the centres of religious activity abroad. Some races, 
like those in Great Britain, who appear to have had a 
special gift for inventing new patterns and combining 
them with a sense of artistic fitness, may have made 
more rapid strides than their neighbours and have 
influenced the development of the style in consequence, 
but that is all that can be said. 

Two special peculiarities of the Italian interlaced- 
work, as compared with that in Great Britain, are the 
ornamenting of the interlaced bands with two incised 




Uj:_^t>; ^i^ >-^5. >, 


1= /\ 



lines instead of one, and the twisting together two 
bands at frequent intervals, thus — 

The latter feature, which is clearly Classical, occurs 
frequently in circular knotwork in this country, showing 
that circular knotwork is of Italian origin. 

The reason why interlaced-work is characteristic of 
early Christian decoration almost throughout the whole 
of Europe, whilst spirals, key-patterns, foliage, etc., 
are confined to particular limited areas, I believe to be 
partly because the number of distinct patterns that can 
be produced from interlaced-work is far greater than 
those which can be got from any other class of 

It is hardly necessary, perhaps, to enlarge much 
upon the subject of the foliage of the early Christian 
period in Great Britain. The scrolls with conventional- 
ised bunches of grapes are no doubt descendants of the 
Classical vine; the involved birds, beasts, etc., being 
a later addition^ of the Bizantino-Barbaro, or Italo- 
Bizantino periods. Foliage is unknown in the Pagan 
Saxon, Scandinavian, or Late-Celtic art, and the only 
other source it could have been derived from is Italian 

We lastly have to consider the parts played in the 
development of early Christian art in Great Britain by 
the Anglo-Saxon and the Scandinavian invaders. 
There does not seem to be much evidence to show that 

^ Or a substitution of later forms for the Cupids, etc., of tiie Classical 


the Saxons were ever gifted with any great capacity 
for ornamental design, although their workmanship 
often reached a high pitch of excellence. In looking 
through the plates of the most recent work on The In- 
dustrial Arts of the Anglo-Saxons^ by the Baron J. de 
Baye, one is struck with the extremely limited range of 
imagination displayed in the design of the patterns. 
Interlaced-work (but of a debased kind) occurs on some 
of the sword-hilts and buckles, the latter evidently bear- 
ing a remarkable affinity to the Merovingian buckles. 
A radiated fibula found at Searby, in Lincolnshire, 
exhibits a diagonal key-pattern similar to that found 
in the Irish MSS. Far the most beautiful specimens 
of Saxon jewellery, however, are the circular brooches 
with cloisonne ornament. The disc-shaped surface of 
these brooches is broken up into little compartments, 
which are filled in with thin slabs of coloured glass, 
garnets, etc. The narrow bands of gold which sepa- 
rate the compartments from each other are zigzagged 
at right angles, or stepped, and it is quite possible that 
the idea of the stepped patterns within circles, which 
occur in the decoration of the Irish MSS. and on the 
circular enamelled bosses on the Irish ecclesiastical 
metalwork, may have been taken from the circular 
cloisonne Saxon brooch. It is only fair, however, to 
mention that circular ornaments of cloisonne enamel, 
with an approximation to a stepped-pattern, are used 
in the decoration of the magnificent Late-Celtic shield 
found in the Thames at Battersea, and now in the 
British Museum. 

It has been suggested that Irish interlaced-work was 
derived from the rude interlaced patterns on the Saxon 
and Merovingian buckles, but this appears to me most 


M. Paul du Chaillu, in his Viking Age, has en- 
deavoured to show that the Anglo-Saxons derived their 
art such as it is from Northern rather than from Central 
or from Western Europe; but his views will not receive 
favour at the hands of the scientific archaeologist who 
relies on hard facts to make good his contentions. 
The forms and ornamental details of the buckles and 
other objects found with Saxon burials in the south 
of England undoubtedly show more affinity with 
Merovingian grave-goods than with anything emana- 
ting from Norway, Sweden, or Denmark. 

Although no trace of Scandinavian influence can be 
detected in the ornamental patterns of the Anglo- 
Saxons — at all events, in the period preceding the 
Viking conquests in the ninth and tenth centuries — 
I am not quite so sure that one of the elements of early 
Christian decorative art in Great Britain may not 
possibly be of Northern origin, namely, the zoo- 
morphic element. I put forward this suggestion with 
the greatest diffidence, and merely as a tentative theory 
until something better can be found to take its place. 

Zoomorphism is not a marked characteristic of Pagan 
Saxon decorative art, and therefore, in order to account 
for the predominance of so-called dragonesque designs 
in the early Irish illuminated MSS., we must fall back 
on one of the following alternatives: (1) that these 
patterns are of native origin, and were invented by 
the Irish ; (2) that like the spirals, they are of Late- 
Celtic origin ; and (3) that they are of Italo-Byzantine 

General Pitt-Rivers and Mr. Henry Balfour have 
given us an insight of the manner in which animal 
forms, by repeated copying, may degenerate into mere 
ornament ; and at one time I thought that early Chris- 


tian zoomorphism might have been the result of a 
process of a reverse nature. It is possible to "see 
snakes " when looking at a piece of interlaced-work 
without necessarily suffering from excess of alcoholism. 
Thus zoomorphic designs might have been evolved 
from interlaced-work by making the bands terminate 
in heads and tails, the limbs following in due course 
later on. Such may have been the process by which 
the Irish illuminator arrived at his zoomorphism, unless 
it can be shown that he got it in some other way. 

Animal forms are comparatively rare in Late-Celtic 
art, and they are not interlaced, so that it is almost 
useless to seek for the original inspiring idea in this 

Birds, beasts, reptiles, and other creatures — often 
used symbolically — are frequently seen in Byzantine 
art, both in the decorative features of churches and in 
the borders of the MSS. If it was thence that the 
early Christian zoomorphs in this country took their 
origin, I fancy the interlacements must have been 
arrived at either by placing the creatures in pairs 
symmetrically facing each other, or by contorting their 
bodies into unnatural attitudes. In the case of beasts 
arranged in pairs, the first step towards interlacement 
is to raise their paws and then to make them cross. 
The beasts may also be placed with their necks crossed; 
their tails may gradually curl round until they pass 
over the body, and may be looped or knotted to fill in 
a blank space ; and in endless other ways the most 
complicated forms of zoomorphic interlaced-work may 
be evolved from simple beginnings. 

Dr. Hans Hildebrand, in his Industrial Arts of 
Scandinavia (p. 50), explains in a most ingenious 
manner how the lion couchant, which so often appears 


in Roman art, forms the basis of the earlier kinds of 
zoomorphic ornament in Scandinavia. The question 
is, did the Irish evolve their zoomorphs independently 
in a similar way from a Classical or Byzantine lion, or 
did they get the idea from the Scandinavians after they 
had so transformed the Roman lion coiichant that all 
resemblance to the original had disappeared? The 
difficulty in settling this point is the absence of accu- 
rately dated specimens of Scandinavian art workman- 
ship. The panels of zoomorphic ornament on some of 
the fibulse of the Later Iron Age, illustrated in Dr. 
Hans Hildebrand's work already referred to (pp. 
58-65), bear a very considerable general resemblance 
to the panels of interlaced beasts in the Irish MSS., 
although the details are worked out differently. The 
whole question turns on the exact date of the Gotland 
brooches. If they can be proved to be earlier than the 
time when zoomorphism first appears in the Irish MSS., 
and if it is possible that the communication between 
Ireland and Gotland can be accounted for by the trade 
in silver objects and bullion existing between this 
country and the East, then there is something to be 
said for the Scandinavian origin of zoomorphism in 
Ireland. I believe, however, that from the evidence 
of the coins found with hoards of silver objects, this 
trade did not begin until about a.d. 800. 

Attention must here be called to two points which 
are common to the zoomorphic and anthropomorphic 
designs of Scandinavia and of Great Britain, namely, 
(i) the introduction of spiral curves to represent con- 
ventionally the folds of the skin where a limb joins 
the body ; and (2) the introduction of figures of men 
grasping birds and beasts, or arranged swastica-wise 
grasping each other's limbs. Here, again, it is not 


easy to decide whether these features were invented 
independently, or whether they were borrowed by the 
Irish from Scandinavia, or by the Scandinavians from 

Whatever may be thought of the possibility of the 
existence of Scandinavian influence on Christian art 
in this country in its earlier phases, there is plenty of 
evidence of the development of an Anglo-Scandinavian 
style in particular districts where the Norse element 
was strong, as in the Isle of Man, and the adjoining 
coasts of Cumberland, Lancashire, and North Wales, 
and in Orkney and Shetland. 

The specially Scandinavian characteristics of the 
sculptured monuments in the districts specified are as 
follows : — 

(i) There is a predominance of patterns formed of chains 
of rings. 

(2) The bands of the interlaced-work have a tendency to 
bifurcate and break off into scroll-like terminations. 

(3) The beasts in the zoomorphic designs have two toes, 
instead of three ; the bodies are covered with scales ; the 
attitude is peculiar, the head being bent back and a crest 
issuing from it with fin-like appendages in places ; and the 
junction of the limbs with the body is conventionally indi- 
cated by spirals. 

(4) Amongst the figure-subjects scenes from the mythic- 
heroic Eddaic poems, such as Sigurd Fafni's bane, Thor 
fishing for the Midgard-worm, Weyland Smith, etc. 

Even in Norman times Scandinavian influence is 
exhibited in the details of the tympana at Hovering- 
ham, and Southwell Minster, Notts, and St. Nicholas, 

The only element in early Christian decorative art 
the origin of which we have not succeeded in running 


to earth in the preceding investigation is the key- 
pattern. I venture to think that this may have been 
suggested by the Greek or Roman fret, and that the 
essentially Celtic character imparted to it was the 
placing of the guiding lines in a diagonal direction 
with regard to the margin, instead of parallel to it. I 
believe the reason for this to be that exactly the same 
setting-out diagram was used both for the interlaced- 
work and the key-pattern. It is often possible to trace 
the origin of key-patterns to the necessities of the 
methods of weaving textile fabrics ; but with regard 
to the ones we are now considering I am inclined to 
think that their beginnings are due to the geometrical 
conditions imposed by the arrangement of the setting- 
out lines. 

In conclusion, I wish to emphasise the fact that the 
beauty and individuality of the ornamental designs 
found in early Christian art in Great Britain are due 
chiefly to the great taste with which the different 
elements are combined and the exquisite finish lavished 
upon them. I cannot see that it in the least detracts 
from the praise due to the orignators of the style if it 
can be shown that the ideas underlying many of the 
patterns were suggested by a pre-existing native style 
or adapted from a foreign one. Interlaced-work, key- 
patterns, spirals, and zoomorphs are to be found 
separately in the decorative art of many races and 
many periods, but nowhere and at no time have these 
different elements been used in combination with such 
consummate skill, as in the early Christian period in 
Great Britain and Ireland. 




HE leading characteristics of Celtic art of the 
Christian period are as follows : — 

(i) The prominence given to the margin or frame within 
which the whole design is enclosed. 

(2) The arrangement of the design within the margin in 
panels, each containing a complete piece of ornament. 

(3) The use of setting-out lines for the ornament, placed 
diagonally with regard to the margin. 

(4) The use of interlaced-work, step-patterns, key-patterns, 
spirals, and zoomorphs in combination. 

(5) The geometrical perfection of all the ornament. 

(6) The superiority of the decorative designs to the figure- 

There are in the world two distinct schools of deco- 
rative art, one which entirely ignores the shape of the 
surface to be ornamented, and the other which allows 
the contour of the margin to influence the whole de- 
sign. Japanese art belongs to the first of these, and 
Celtic art to the second. In the Irish illuminated 



MSS. the rectangular shape of the page determines 
the setting out of the design, which is universally 
enclosed within a rectangular margin composed of 
lines of various thicknesses, or within an ornamental 
panelled frame. The only exception is in the case of 
the initial pages of the Four Gospels, where the margin 
is incomplete, so as to allow the extremities of the 
letters to project more nearly to the edge of the page. 
This prominence given to the margin often greatly 
influences the designs within it, more especially the 
key-patterns with diagonal setting-out lines. In sculp- 
tured stonework either roll-mouldings or flat bands 
form the margin, and in metalwork the margins are 
raised and the panels sunk. 

The panels within the margin are generally rect- 
angular, but sometimes they are circular, annular, 
segmental, triangular, etc. The ornament in adjoining 
panels is seldom of a similar kind, and the patterns 
are often arranged on the principle of chequerwork, 
so that if there is a panel of interlaced-work at the 
left-hand upper corner of the page of a MS., and a 
panel of key-pattern at the left-hand lower corner, the 
order will be reversed on the opposite side of the page, 
and the key-pattern will be at the right-hand upper 
corner and the interlaced-work at the right-hand lower 

The diagonal setting-out lines are chiefly confined to 
the key-patterns, and, as we shall see subsequently, 
are the origin of the peculiar form of Celtic key-pattern 
which was developed from the Greek fret. 

The various motives that have been specified — 
namely, interlaced-work, step-patterns, key-patterns, 
spirals, and zoomorphs — are not always found in com- 
bination, except in the MSS., sculptured stones, and 


metalwork of the best period. The step-patterns are, 
as a rule, only found in the early MSS. and on the 
enamelled settings of metalwork. Foliage is a dis- 
tinctly non-Celtic element, and wherever it occurs it is 
a proof of Anglian influence from Northumbria. As 
the decadence of Celtic art set in the spirals dis- 
appeared first, and then the key-patterns, leaving only 
interlaced-work and zoomorphs, which survived even 
after the Norman conquest. Key-patterns survived in 
a debased form in the architectual details of the 
churches of the twelfth century in Ireland, but not in 
Scotland or Wales. 

By the geometrical perfection of the Celtic ornament 
is meant that there are hardly ever any mistakes in the 
setting-out and complete execution of the designs. 
Thus in the interlaced-work every cord laps under and 
over with unfailing regularity (never over two or under 
two), and all the cords are joined up so as not to leave 
any loose ends. All the details of the spiral-work are 
executed with the minutest care, and there is never 
a broken line or pseudo-spiral. In the zoomorphic 
designs the beasts are all provided with the proper 
number of limbs and are complete in every respect 
down to the smallest detail. 

The inferiority of the figure drawing in Christian 
Celtic art to the ornament will be dealt with sub- 
sequently in its proper place. 

We will now proceed to examine in detail the 
different motives made use of in the Celtic art of the 
Christian period in Great Britain. 



The interlaced ornament used in Celtic art may be 
divided into the following classes : — 

(i) Regular plaitwork, without any breaks. 

(2) Broken plaitwork, with breaks made in an irregular 


(3) Knotwork. 

(4) Circular knotwork. 

(5) Triangular knotwork. 

(6) Ringwork or chainwork. 

Interlaced-work is the predominant motive of the 
Celtic style of the Christian period. It lasted longer 
in time than any other motive, and its geographical 
distribution extends over a larger area. It is very 
seldom that one motive is used by itself for the decora- 
tion of a stone monument, metal object, or page of a 
MS.; but where this is the case the motive chosen is 
invariably interlaced-work, and not a key-pattern, spiral, 
or zoomorph. As instances of sculptured monuments 
decorated entirely with interlaced-work we have the 
cross at Neuadd Siarman, Brecknockshire, and the 
cross-shaft at St. Neot, Cornwall. 

The evolution of knotwork from plaitwork cannot 
better be studied anywhere than in the decoration of 
the Welsh crosses. Let us now endeavour to trace the 
various stages in the process by which the higher 
forms of Celtic interlaced work were arrived at. 

In Egyptian, Greek, and Roman decorative art the 
only kind of interlaced-work is the plait, without any 
modification whatever ; and the man who discovered 
how to devise new patterns from a simple plait by 
making what I term breaks laid the foundation of 
all the wonderfully complicated and truly bewildering 


forms of interlaced ornament found in such a master- 
piece of the art of illumination as the Book of Kells 
in the library of Trinity College, Dublin. Although 
we do not know idio made this discovery of how to 
make breaks in a plait, we know pretty nearly isohen 
it was made. In the decoration of the mosaic pave- 
ments in Great Britain belonging to the period of the 
Roman occupation, no instance, as far as I can ascer- 
tain, exists of the introduction of a break in a plait ; 
nor is there any break in the plaitwork on the marble 
screen and the capitals of the columns of the ciborium 
in the Church of San Clemente at Rome (which are 
dated by R. Cattaneo^ between a.d. 514 and 523). In 
the eighth century, however, there are several examples 
with well-authenticated dates of the use of true knot- 
work (as distinguished from plaitwork) in the decoration 
of churches in Italy ; namely, on the ciborium of San 
Giorgio at Valpolicella'^ (a.d. 712); on the Baptistery 
of Calistus at Cividale'^ (a.d. 737); and on the jambs 
of the doorway of the Chapel of San Zeno in the 
Church of San Prassede at Rome^ (a.d. 772-795). 

It would appear, then, that the transition from plait- 
work to knotwork took place between the Lombard 
conquest of Italy under Alboin in a.d. 563, and the 
extinction of the Lombard monarchy by Charlemagne 
in A.D. 774; possibly during the reigns of Luitprand 
(a.d. 712-736) and Rachis (a.d. 744); for the name of 
the former king is mentioned in the inscriptions on 
the Baptistery at Cividale and the ciborium of San 
Giorgio at Valpolicella, and the latter on the altar at 

^ L' Architettura in Italia^ pp. 29 and 31. 

2 Ihid,, p. 80. •' Ihid., p. 87. 

■* Archceologia , vol. xl. , p. 191. 

IXz2iIi22iI ICiTlZ^rAl TC W2l^ ^' ^ K' 

H fctq 

■^ o 

O -1 


I now propose to explain how plaitwork is set out, 
and the method of making breaks in it. When it is 
required to fill in a rectangular panel with a plait the 
four sides of the panel are divided up into equal parts 
(except at the ends, where half a 
division is left), and the points 
thus found are joined, so as to 
form a network of diagonal lines. 
The plait is then drawn over these 
lines, in the manner shown on 
the accompanying diagram. The 
setting-out lines ought really to 
be double so as to define the 
width of the band composing 
the plait, but they are drawn 
single on the diagram in order 
to simplify the explanation. 

If now we desire to make a 
break in the plait any two of the 
cords are cut asunder at the point 
where they cross each other, leaving four loose ends 
A, B, C, D. To make a break the loose ends are joined 
together in pairs. This can be done in two ways only: 
(i) A can be joined to C and D to B, forming a vertical 

Regular plaitwork without 
any break 

AD \ : 

c B ,' \ 

/ \ /C B\ 


C ,- 



•» B 

Method of making- breaks in plaitwork 

break ; or (2) A can be joined to D and C to B, forming 
a horizontal break. The decorative effect of the plait is 
thus entirely altered by running two of the meshes 



between the cords into one. By continuing the process 
all the knots most commonly used in Celtic decorative 
art may be derived from a simple plait. 

Let us proceed to trace the process of the evolution 
of knotwork out of plaitwork by actual instances taken 
from the Welsh crosses. We have, to start with, good 
examples of plaits of four, six, and ten cords ^ without 
any breaks at Nevern, Pembrokeshire ; and Llantwit 

Regular plaitwork, with one 

vertical break and one 

horizontal break 

Six-cord plait, with 

horizontal breaks at 

regukir interviils^ 

Major, and Margam, Glamorganshire. Next, plaits 
with a single break only are to be seen at Carew, 
Pembrokeshire, and Llantwit Major, Glamorganshire ; 
then plaits with several breaks, made quite regardless 
of symmetry or order, at Golden Grove, Carmarthen- 
shire ; and, lastly, breaks made at regular intervals, 

^ Plaits of an uneven number of cords are seldom used, because they 
produce lopsided patterns. 

■^ This occurs on the second panel of the cross at Llanbadarn Favvr. 

Cross-shaft at Llantwit Major 

(No. 5), Glamorg-anshlre. 

Eight-cord plait, with cruciform 


Scale iK linear 

Cross-shaft at Golden Grove, with panels 
of irregular broken plaitwork 

Scale iV linear 



at Neuadd Siarman, Brecknockshire. When the breaks 
are made symmetrically at regular intervals, and brought 
sufficiently near together, the plait ceases to be the 





Eig-ht-cord plaits, with cruciform breaks 

most prominent feature in the design, and in its 
place we get a pattern composed entirely of what 
(for want of a better name) are called knots. On 


some of the Welsh crosses (as at Carew and Nevern, 
Pembrokeshire), however, the breaks are made with 
sufficient regularity and proximity to produce knots, 

Eight-cord plaits, with cruciform breaks 

and yet the knots themselves are not symmetrically 
placed. The result is a class of interlaced-work, inter- 
mediate between plaitwork with irregular breaks and 



Six-cord plait, with cruciform 
(Occurring- at Llanbadarn 

Ten-cord plait, with cruciform 
(Occurring- at St. Neuadd 



Knots derived from a three-cord plait 


knotwork. The same kind of thing is to be seen on 
the crosses at Coppleston, Devonshire ; and St. Neot, 

If two horizontal breaks and two vertical breaks are 
made next to each other in a plait, a space in the shape 
of a cross is produced. A large number of the inter- 
laced patterns used in Celtic decorative art are derived 
from a plait by making cruciform breaks at regular 
intervals. There are examples of this in Wales, at 
Neuadd Siarman, Brecknockshire ; Llanbadarn Fawr, 
Cardiganshire ; and Llantwit Major, Glamorganshire. 
It is not unlikely that symbolism had something to do 
with the frequent use of the cruciform break. 

There are eight elementary knots which form the 
basis of nearly all the interlaced patterns in Celtic 
decorative art, with the exception of those already 
described. Two of the elementary knots are derived 
from a three-cord plait, and the remaining six from a 
four-cord plait. 


Knot No. I Knot No. 2 

Knot No. 1 is derived from a three-cord plait by 
making horizontal breaks on one side of the plait only, 
and No. 2 by making horizontal breaks alternately on 
one side and the other. 

Knot No. 3 is derived from a four-cord plait by 
making horizontal breaks in the middle of the plait. 


Knot No. 4 is derived from No. 3 by making a 
horizontal break at A ; and No. 5 from No. 4 by 
making a vertical break at B and C. 

Knot No. 6 is derived from a four-cord plait by 
making horizontal breaks in the middle of the plait, 
in the same way as in the case of knot No. 3, but 
closer together. 


Knot No. 4 Knot No. 3 Knot No. 5 

'0. m m 

Knot No. 7 Knot No. 6 Knot No. 8 

Knot No. 7 is derived from No. 6 by making a 
vertical break at B ; and No. 8 from No. 6 by making 
vertical breaks at B and C. 

If a series of knots repeated in a single row can be 
derived from a plait of n bands, a series of the same 
knots repeated in a double row can be derived from a 
plait of 2n bands. Thus a pattern composed of knot 



No. I arranged in a double row would be derived from 
a plait of six cords. 

Knots like Nos. 3 and 4, which are longer than they 
are broad, can be placed either horizontally or verti- 
cally. Thus No. 3 placed with its longer axis vertical 
can be derived from a four-cord plait, but if placed 
horizontally it would be derived from a six -cord 


Method of deriving- Knots Nos. 3 and 6 from a four-cord plait 

Knot No. 2 does not occur on the Welsh crosses, 
and No. i only in a double row, as at Neuadd Siarman, 
Brecknockshire. This pattern is derived from a six- 
cord plait by making horizontal breaks in the two 
edges of the plait, and vertical breaks in the middle, 
the stages being shown on the annexed diagram. 

Knot No. 3, in a single row placed with its longer 
axis vertical, occurs at Llandough, Glamorganshire, 
and, in a single row placed the other way, at Margam, 

Examples of the two knots, Nos. 4 and 5, which are 



Knots Nos. 4, 5, 7, and 8, derived from a four-cord plait 




Knot No. I, derived from either a three-cord or a six-cord plait 


Knots 3 and 4, derived from a six-cord plait 

Evolution of Knot No. i from a six-cord plait 


derived from No. 3, are to be seen at Baglan, Glamor- 
ganshire, and Penally, Pembrokeshire. 

Knot No. 6, in a single row, occurs at Llantwit 
Major, Glamorganshire, and its second derivative, 
No. 8, at Llantwit Major, and also at Neuadd Siarman, 
Brecknockshire. Its first derivative. No. 7, is only 
used in a double row on the Welsh crosses, as at 
Silian and Maes Mynach, Cardiganshire, and at 
Penally, Pembrokeshire, where the knots have an 
extra spiral twist. The direction of the twist of the 
spirally bent cord is the same in both the right-hand 
and left-hand vertical row of knots, although the 
positions of the knots are different. The more usual 
arrangement is to make the cords twist in opposite 
directions, as on the annexed diagram, in which the 
evolution of the pattern is shown. (Page 271.) 

The clearest proof that the spiral knot No. 7 was 
developed from plaitwork in the manner explained 
is that on stones at Llangenydd, Glamorganshire ; 
Whithorn, Wigtownshire ; Abercorn, Linlithgowshire; 
and Aycliffe, Co. Durham ; the successive stages of 
development can be easily traced. 

I have coined the term circular knotwork to describe 
a particular class of interlaced-work, in which the 
circular curves made by the cords give the pattern its 
distinctive appearance. The best example of circular 
knotwork in any of the Hiberno-Saxon MSS. occurs 
on one of the ornamental cross-pages of the Book of 
Durrow.^ Circular knotwork is not used in the deco- 
ration of the Irish ecclesiastical metalwork, probably 
because it is only suitable for application to larger 

^ J. R. Allen and J. Anderson's Early Christian Momunents of Scot- 
land^ p. Ixxviii. ; J. A. Bruun's Jlliimiiiated Manuscripts of the Middle 
Ages, pt. I, "Celtic JMSS.," p. 8. 


surfaces than are to be found on comparatively small 
metal objects. Circular knotwork is characteristic of 

Evolution of Knot No. 7 from an eight-cord plait 

the Irish and Scottish sculptured monuments of the 
best period ; it is unknown in Cornwall and the Isle of 
Man and there is only one instance of its occurrence in 



Wales. Very good examples of circular knotwork 
may be seen on sculptured monuments in Ireland ^ at 
Kells, Co. Meath ; Monasterboice and Termonfechin, 
Co. Louth ; Boho, Co. Fermanagh ; Kilfenora, Co. 
Clare ; and Drumcliff, Co. Sligo ; and in Scotland^ at 
Collieburn, Sutherland (now in the Dunrobin Museum); 
Tarbet (now at Invergordon Castle), Brodie, Elgin- 
shire ; Nigg, Ross-shire ; Aberlemno, Monifieth (now 

Evolution of Knot No. 7 iVoni an eight-cord plait 

in the Edinburgh Museum), and Eassie, Forfarshire ; 
and Rossie Priory and St. Madoes, Perthshire. 

The most common kinds of circular knotwork appear 
to have been evolved in the following manner. It has 
already been shown how knot No. 3 can be derived 
from a four-cord plait by making a series of horizontal 
breaks at regular intervals, leaving two crossing-points 
of the cords between each break ; and how knot No. 4 

^ H. O'Neill's Crosses of Ancient Ireland. 

^ Allen and Anderson's Early Christian Monuments of Scotland. 


can again be derived from knot No. 3 by making a 
horizontal break at the point A. 

Knot No. 3 

Knot No. 4 

Now if a pair of knots like No. 4 be placed opposite 
each other thus — 

and repeated in a vertical row, we get the pattern 
shown below. 



By making pointed ends to the loops forming the 
knots and "sweetening" the curves of the bands 
between each knot the appearance of the whole is 
changed, and its development from the plait disguised. 
Almost all geometrical ornament is capable of convey- 
ing several different impressions to the mind according 
to the way it is observed by the eye for the time being, 
and the intellectual pleasure which a pattern gives is 
most probably dependent on the infinite variety of 

Sections of pattern shown on p. 273 

these kaleidoscopic changes. Taking this pattern for 
example, if the attention is concentrated upon the 
portions of the pattern between each of the points 
where the bands cross in the centre, it will seem as if 

Knotwork from Ramsbury, Wilts, and Nig-g, Ross-shire 

the whole was formed of repetitions of knot No. 4 ; 
but if the attention be now directed towards the por- 
tions lying between the middle points of each of the 


Photograph supplied by the Rev. 'J. J/. -yoass, LL.D., Honorayy Curator 


knots, the pattern will appear to consist entirely of 
circular curves with two diameters crossing each other 

When the circular knot thus obtained is repeated 
in a double row we get a comparatively simple pattern, 
in which the circular curves assume much greater 

More complicated forms of circular knots can be 
derived from the elementary circular knot by com- 
bining it with a circular ring, either a larger one 
enclosing the four loops in the middle entirely, or a 
smaller one interlaced through the loops thus : — 

Further variations"can again be produced from these 
by severing the bands in places, and joining parts of 
the loops to the rings on the same principle that breaks 
can be made in a plait. 

The connection between the different knots will at 
once become clear if they are drawn on separate pieces 
of tracing paper and placed one over the other. 



Another kind of circular knotwork is formed by 
enclosing the simpler sort of knots derived directly 
from plaitwork within a circular band, which crosses 
over in one or two places and turns inwards to form 
the enclosed knots. 

Circular knotwork from Tarbet, Ross-shire 

The illustrations of the different kinds of circular 
knotwork from actual examples show the process of 

Circular knotwork from Monasterboice, Co. Louth 



By the term triangular knotwork is meant interlaced 
patterns, the setting-out lines of which form triangles 
only or triangles and lozenges. The patterns are made 
by distorting the simple knots derived from plaitwork, 
so as to adapt them to the triangular shape. This 
species of knotwork is very seldom seen except in a 
few of the Hiberno-Saxon MSS. and on some of the 
sculptured stones of Ireland and Scotland. The best 

Triangular knotwork from Ulbster, Caithness 

examples are at Kilfenora, Co. Clare ; Ulbster (now at 
Thurso Castle), Sutherlandshire ; and Dunfallandy, 

Under the head of ringwork and chainwork are in- 
cluded all patterns composed of circular, oval, and 
looped rings interlaced symmetrically round a centre, 
or arranged so as to form a long chain. Patterns of 
this kind are not found in the best Celtic work, and 
when they occur it is generally an indication either of 
Scandinavian influence or of the style being debased. 



A certain number of modifications of the interlaced- 
work already described are produced by adapting the 
patterns so that they will fit into circular or annular 
spaces. Instances of this may be seen on the erect 

Triangular knotwork from Dunfallandy, Perthshire 

cross-slabs at Hilton of Cadboll (now at Invergordon 
Castle) and Nigg, Ross-shire ; Glamis, Forfarshire ; 
and Rossie Priory, Perthshire; and on the Lough Erne 
and Monymusk Reliquaries. 


A step-pattern is one which is formed of straight 
lines bent backwards and forwards at right angles so 
as to resemble a flight of steps. The lines are often 


arranged symmetrically round a centre, so as to make 
cruciform and swastika designs, and the different parts 
are also generally shaded alternately black and white 
on the principle of chequerwork. Step-patterns hardly 
ever occur in Christian Celtic art except on the ena- 
melled bosses of metalwork and in a few of the 
illuminated MSS., such as the Lindisfarne Gospels, 
the St. Gall Gospels (Codex No. 51), the Gospels of 
MacRegol, the Book of Kells, and the Book of 
Durrow. The step-patterns in the MSS. so nearly 
resemble those on the enamelled bosses on the Ardagh 
Chalice, the Tara Brooch, and the Cross of Cong, that 
there can be but little doubt the illuminators copied 
their designs from the enamels. In the Pagan Celtic 
enamels the ornament is nearly always curvilinear ; but 
in the Christian Celtic enamels it is rectilinear, the 
arrangement of the cloisons being very similar to that 
on the Anglo-Saxon disc brooches incrusted with small 
slabs of garnet, glass, etc. Instances have already 
been given in a previous chapter of the use of step- 
patterns by the Pagan Celts on the engraved woodwork 
from the Glastonbury Marsh Village (p. 161). The 
only instances I have met with of step-patterns on the 
sculptured stones of the early Christian period in this 
country are at Brad ford-on- A von, Wilts ; and Dysert 
O'Dea, Co. Clare. 


The term key-pattern is used to describe a particular 
kind of rectilinear ornament which bears a certain 
amount of resemblance to the perforations in a key to 
allow it to pass the wards of a lock. The best-known 
key-pattern is the Greek fret. This is composed of 
what may be appropriately called straight-line spirals ; 


(i) Aberlady, Haddingtonshire (2) Abercorn, Linlithg-owshire 

(3) St. Andrews, Fifeshire (4) Collieburn, Sutherlandshire 


that is to say, straight lines (or, to speak more accu- 
rately, narrow straight bars) bent round into a series 
of right angles in the same direction. The space 
between the lines (or narrow bars) is generally about 
the same width as that of the line itself. 

The key-patterns used in Christian Celtic art may be 
classified as follows : — 

(i) Square key-patterns^ in which the lines run horizontally 
and vertically parallel to the margins. 

(2) Diag07ial key-patterjis ^ in which the lines run vertically 
parallel to the right and left margins, and diagonally in two 
directions at an angle of 45° to the margins. 

(3) Diaper key-patterns^ in which the lines run horizontally 
and vertically parallel to the margins, and diagonally in two 
directions at an angle of 45° to the margins. 

The essential difference between the key-patterns 
used by the Egyptians, Greeks, and Romans and those 
used by the Christian Celts consists in the introduction 
of diagonal lines by the latter. Square key-patterns 
{i.e, those of the Greek fret type) were very seldom 
used in Christian Celtic art. There is, however, a 
very good example on one of the crosses at Penmon, 
Anglesey. The first step in the evolution of the Celtic 
key-pattern was to turn the Greek fret round through 
an angle of 45° so as to make the lines run diagonally 
with regard to the margins instead of parallel to them. 
Key-patterns in this stage of development are to be 
seen in the Lindisfarne Gospels, the St. Gall Gospels 
(Codex No. 51), and an Anglian cross-shaft from 
Aberlady, Haddingtonshire, now at Carlowrie Castle, 
near Kirkliston, Midlothian. It will be observed, how- 
ever, that the result of changing a square key-pattern 
into a diagonal one is to leave a series of unornamented 


(i) Rosemarkie, Ross-shire 
(3) Gattonside, Roxburghshire 

(2) Farr, Sutherlandshire 
(4) Nig-g, Ross-shire 


triangles all round the edge (p. 280). When these tri- 
angles are filled in by bending the ends of the diagonal 
lines round through an 
angle of 45°, so as to run 
parallel to the margins, 
we get such a character- 
istically Celtic key-pattern 
as the one on the great 
cross-shaft at St. Andrews, 
Fifeshire. Lastly, when 
the opposite ends of the 
diagonal lines in the mid- 
dle of the panel are bent 
round in a similar manner, 
the most typical of all the 
Celtic key - patterns is 
arrived at, of which there 
is a very good example 
on the erect cross-slab at 
Farr, Sutherlandshire. 

The filling in of the 
sharp corners made by 
the lines inclined to each 
other at an angle of 45°, 
with small black triangles 
(if in a MS.) or with 
sunk triangles (if on a 
sculptured stone) gives a 
decorative finish to the 
pattern, and still further 
adds to its distinctively 
Celtic character. 

Next to interlaced-work 
the key -pattern is the 







Shaft of Cross of Eiudon, at Golden 
Grove, Carmarthenshire 



most common motive made use of in the decorative art 
of the Christian Celtic period. It occurs in nearly all 
the Hiberno-Saxon illuminated MSS. and on a large 
proportion of the sculptured monuments in Ireland, 
Scotland, and Wales. Key-patterns and interlaced- 

I 1 

Detail of ornament on erect cross-slab at Rosemarkie, Ross 


work in combination, but without any other decorative 
motive, may be seen on the Welsh crosses at Carew 
and Nevern, Pembrokeshire ; Golden Grove, Car- 
marthenshire ; and Llantwit Major and Margam, 
Glamorganshire. On the metalwork of the period 
key-patterns seldom occur, except on the bronze bells, 
on a strap-buckle from Islandbridge, near Dublin, and 
on the Crucifixion plaque of repousse bronze from 
Athlone, now in the Dublin Museum. 



The spiral is the only decorative motive used in 
Christian Celtic art that can be proved to have been 
borrowed from the Pagan Celtic art of the preceding 
period. Although spiral ornament appears to be so 
complicated when it is completed, the geometrical pro- 
cess of setting it out is simplicity itself. All that it is 
necessary to do is to fill in the surface to be decorated 

Methods of connecting- spirals 

with circles of any size, leaving about the same distance 
between each ; then connect the circle with S- or C- 
shaped curves ; and, lastly, fill in the circles with 
spirals working from the tangent points, where the 
S or C curves touch the circles, inwards to the centre. 
As the size of the circles is a matter of no importance, 
a surface of irregular shape may be covered with spiral 
ornament just as easily as one of symmetrical shape. 
In the flamboyant ornament of the Pagan Celtic period 
we have the same S- and C-shaped curves, but the 
circles were occupied either by a disc of enamel (as on 



the bronze shield from the Thames), or by raised 
conchoids (as on the gold necklet from Limavady, Co. 
Londonderry). In the spiral ornament of the Christian 
Celtic period closely coiled spirals like those of the 
Bronze Age were substituted for the discs of enamel 
or raised conchoids. The background of the spirals, 
however, retained several of the prominent features of 
the repousse metalwork, the effect of the light shining 




::;\BCTi i 



Tree key-pattern, Meigle, Perthshire 

on the raised trumpet-shaped expansions of the S and 
C curves being imitated in black and white or coloured 
by almond-like dots. In the later and less refined 
spiral ornament of the Christian period this back- 
ground disappears altogether, and the spirals are made 
all the same size and placed close together. 

As the spiral was the earliest decorative motive in 
Christian Celtic art, so it was also the first to disappear, 
and its disappearance marks the decadence of the style. 
We have in a previous chapter traced the spiral motive 


from the Pagan metalwork through the enamelled 
disc ornaments of the bronze bowls of the Transition 
period to the illuminated MSS. of Christian times. 

Spiral ornament, with key-pattern border, from the Book of Kells 

Spiral ornament in its best form is to be found in the 
following MSS. :— 

The Lindisfarne Gospels 
The Book of Kells . 
The Gospels of St. Chad 
The Book of Durrow 
The Book of Armagh 
The Gospels of Willibrod 
The Gospels of St. Gall 
The Gospels of MacRegol 
The Gospels of Stockholm 
The Vespasian A. i. Psalter 

In metalwork spiral ornament is less common than 
in the MSS., there being good examples on the Ardagh 





> ) 



J ) 






or 9th cer 






8th century. 


Chalice, the Tara Brooch, the Hunterston Brooch, the 
Monymusk Reliquary, and the Athlone Crucifixion 

Spiral ornament of the best kind is found on the 
sculptured stone monuments only in Ireland and 
Scotland.! In Wales, Cornwall, and the Isle of Man 
spiral ornament is extremely rare, and when it does 
occur it is of debased character. Typical examples 
of spiral ornament may be seen on the sculptured 

Method of connecting- spirals 

monuments in Ireland at Kells, Co. Meath ; Monaster- 
boice, Co. Louth ; Clonmacnois, King's Co. ; and 
Kilklispeen, Co. Kilkenny ; and in Scotland at Nigg, 
Shandwick, Hilton of Cadboll (now at Invergordon 
Castle), all in Ross-shire ; the Maiden Stone, Aber- 
deenshire; St. Vigeans, Forfarshire; Meigle and Dun- 
fallandy, Perthshire; and Ardchattan, Argyllshire. 

Judging from the evidence afforded by the dated 
specimens, the best kind of spiral ornament seems to 
have disappeared entirely from Christian Celtic art after 
the first quarter of the tenth century. 


Animal forms are used in Celtic art of the Christian 
period in three different ways, namely, (i) pictorially, 

^ Chiefly in the Pictish districts of the north-east of Scotland. 





"*^4^^K^^f€^S^I?#^'*-- -"m^im^ 




c c c 


(2) symbolically, and (3) decoratively. As cases of the 
first kind of treatment we have the hunting scenes,^ 
battle scenes,'-^ men driving in chariots'^ drawn by 
horses, groups of animals,^ etc., on the erect cross- 
slabs of Scotland and on the bases of some of the 
Irish and Welsh crosses. Although these subjects 
may have some symbolism behind them, yet all the 
living creatures represented are treated realistically, 
and not conventionally. As cases of the second kind 
of treatment we have the Symbols of the Four Evan- 
gelists, which, although consisting of the figures of 
a man, a lion, a bull, and an eagle, are generally 
highly conventionalised. Lastly, we have the decora- 
tive use of animal forms, where the zoological species 
of the creatures represented becomes of so little im- 
portance that it is altogether ignored. The creatures 
can certainly be divided into beasts, birds, fishes, and 
reptiles ; but the artist has taken such liberties with 
the shapes of their bodies, limbs, heads, tails, and 
other details, that he would be a bold man who would 
say of any one of the beasts whether it was intended 
for a lion, a tiger, a dog, a wolf, or a bear. The 
quadruped most in favour with the Christian Celtic 
artist may, as has been already suggested, have been 
degraded by successive copying from the Classical lion. 
Anyway, it has a head something like that of a dog, 
with pointed ears, an attenuated body, four legs termi- 
nating in paws with claws, and a long tail. The head 

^ As on the erect cross-slab at Hilton of Cadboll (now at Inverg-ordon 
Castle), Ross-shire. 

'^ As on the erect cross-slab at Aberlemno, Forfarshire. 

"^ As on the base of the cross in the churchyard of Kells, Co. Meath. 

^ As on the erect cross-slabs at Shandwick, Ross-shire ; emd St. 
Vig-eans, Forfarshire ; and on the base of the cross at Castledermot, 
Co. Kildare. 



and the paws are never misrepresented for decorative 
purposes, but the body, limbs, ears, and tail may be 
extended to any given length or bent in any desired 
direction. The beasts and other creatures are generally 
shown in profile, and only rarely in plan. 

In the simplest kind of zoomorphic ornament a single 
beast is used to fill a panel, the different attitudes being 
as follows : — 

(i) With the head looking forwards. 

(2) With the head bent backwards. 

(3) W^ith the head bent backwards, biting the middle of 

the body. 

(4) With the head bent backwards, biting the end of 

the tail. 

(5) With the tail curled up over the back. 

(6) With the tail curled up under the belly. 

If the beasts are in pairs, they may be placed in the 
following positions : — 

(i) Symmetrically facing towards each other, or face 
to face. 

(2) Symmetrically facing away from each other. 

(3) In a horizontal row one in front of the other. 

(4) In a vertical row one below the other. 

When there are three or four beasts, besides being 
arranged in rows, they may be placed after the fashion 
of the triskele or the swastika round a centre. 

Interlaced zoomorphic ornament can be made with a 
single beast by extending the length of its tail and ear, 
and forming them into knots at intervals, crossing 
over the body and limbs where necessary. Sometimes 
the tail alone is knotted. In this sort of ornament the 
shape of the beast is seen distinctly and the knots 
occupy the background. A more complicated design 


can be made from a single beast by twisting its body 
and limbs into knots as well as the ears and tail. 

The panels of zoomorphic ornament in Christian 
Celtic art are, however, usually composed of two or 
more beasts placed symmetrically with regard to each 
other and having their bodies and limbs crossed over 
and interlaced. The ears and tails may also be ex- 
tended and formed into knots in combination with the 
bodies and limbs. The designs thus produced will 
be seen to consist apparently of two sets of bands 
crossing each other diagonally, the wide bands being 
the bodies of the beasts and the narrow bands, the 
limbs, tails, and ears. The bands are nearly straight, 
or if bent at all only gently curved. 

When the beasts are not placed in opposite sym- 
metrical positions, but in horizontal rows one in front 
of the other, or in vertical rows one below the other, 
the bodies are often bent round spirally in one direction 
or twisted into S-shaped spirals in two directions. A 
favourite device with the Celtic artist was to make the 
beasts bite their own bodies, limbs, or tails, or the 
body, limbs, or tail of the beast immediately in front of it. 

The zoomorphic designs composed of birds were 
arranged on the same principles as those composed of 

Reptiles or serpentine creatures with bodies of 
nearly the same width throughout were converted into 
interlaced zoomorphic ornament by twisting, plaiting, 
looping, or knotting the bodies together. This class 
of ornament is, in fact, the ordinary interlaced patterns 
derived from the plait, with heads added at one end 
and tails at the other. 

A very ingenious zoomorphic design is made by 
filling in a long narrow panel with the body of a 


serpentine creature undulating from side to side. The 
head of the creature is at the top of the panel, and the 
body remains about the same width until it reaches the 
bottom of the panel, where its width is greatly reduced 
and its direction reversed. On its return journey it 
makes a series of Stafford knots, which fill in the 
spaces between the undulations of the body and the 
sides of the panel, and the end of the tail is finally 
received into the mouth of the reptile.^ 

There are two kinds of zoomorphic designs which 
are peculiar to the MSS. of the period, namely, initial 
letters made in the form of a bird or beast, and the 
incomplete frames round the initial pages of the 
Gospels terminating in a beast's head at one end and a 
fish-like tail at the other. The only thing of a similar 
kind which occurs on the sculptured monuments is the 
zoomorphic margin round some of the erect cross- 
slabs of the east of Scotland.'-^ The margin is formed 
by two beasts, the heads of which appear at the top 
facing each other and the tails at the bottom. 

Zoomorphs are found throughout the whole range 
of Christian Celtic art ; they form an important feature 
in the decoration of nearly all the Hiberno- Saxon 
illuminated MSS. ; they are particularly characteristic 
of the Irish ecclesiastical metalwork ; and they are of 
frequent occurrence on the sculptured monuments of 
Ireland and Scotland. On the crosses of Wales and 
Cornwall zoomorphs are comparatively rare. Some of 

^ Instances of this occur at Lanherne and Sancreed, Cornwall. 

2 At Cossins and Monifieth, Forfarshire ; and Meigle, Dunfallandy ; 
and St. Madoes, Perthshire, The arched top of the frame round the 
miniature of Christ seized by the Jews, in the Book of Kells, is treated 
in exactly the same way as the pedimented tops of the erect cross-slabs. 
In the second table of Eusebian Canons, in the Book of Kells, the head 
and arms of Christ are placed between the two beasts' heads. 

■- ~v-. ,:iy/v 






From pliolografhs of the ca^t in the Si/atrc a/id .Irt Museinii, lldhilntrgh, 
supplied by Mr. l-'aliaiuc, Curator 


the best instances of zoomorphic designs in the MSS. 
are to be seen in the cross-pages of the Book of Durrow, 
the Gospels of Lindisfarne, the Book of Kells, the 
Gospels of St. Chad, and the St. Gall Gospels (Codex 
No. 51); in metalwork on the Ardagh Chalice, the 
Tara Brooch, the Hunterston Brooch, the Shrine of 
the Bell of St. Patrick's Will, the Cross of Cong, and 
the Shrine of St. Manchan ; and on sculptured monu- 
ments at Termonfechin, Co. Louth ; Kells, Co. Meath ; 
Tihilly, King's Co. ; Dysert O'Dea, Co. Clare; Nigg, 
Ross-shire ; Aberlemno and Invergowrie, Forfarshire ; 
St. Madoes, Perthshire; Penally, Pembrokeshire; and 
Sancreed and Lanherne, Cornwall. 

Sometimes key-patterns and spirals are converted 
into zoomorphic designs by the addition of animals' 
heads, as at Penmon, Anglesey ; and Termonfechin, 
Co. Louth. The centres of spirals are also often made 
zoomorphic, as in the Gospels of Lindisfarne, on the 
cross at Kilklispeen, Co. Kilkenny, and on an erect 
cross-slab at St. Vigeans, Forfarshire. 

Probably the most wonderful tour de force in the way 
of zoomorphic sculpture is a pair of panels on the 
erect cross-slab at Nigg, Ross-shire. Each panel is 
ornamented with a series of raised bosses arranged 
symmetrically. The whole of the convex surfaces of 
the bosses is covered with intricate knotwork, and the 
background is composed of serpents, the tails of which 
coil spirally round the bases of the bosses, and in each 
case enter the circumference at three points to form the 
interlaced-work on the boss. After innumerable cross- 
ings under and over, the tails again diverge at three 
other points round the base of the boss, and finally 
terminate in small spirals in different parts of the back- 



Under the above heading are classed all designs in 
which the complete figure of a man, or portions of a 
man are used for purposes of decoration. Human 
heads occur in metalwork in the decoration of the Tara 
Brooch 1 and in sculptured stonework on the cross of 
Muiredach, at Monasterboice, Co. Louth, and on a 
cross-head from the crannog at Drumgay Loch. 

The most remarkable instances of the decorative use 
of the complete figures of men in the illuminated MSS. 
are to be found in the Book of Kells. The figures 
are generally arranged in pairs facing each other, in 
groups of three triskele fashion, and in nearly all cases 
the attitudes are extremely uncomfortable with the 
knees drawn up close against the stomach. The limbs 
of the different figures are crossed over and interlaced, 
as in zoomorphic ornament, and the hands are shown 
grasping either the limbs, hair, or beard of one of the 
other figures. Sometimes the human figures are com- 
bined with figures of birds or beasts. 

We have already referred to the incomplete frames 
of the initial pages of the Gospels with zoomorphic 
terminations. In the '^ Nativitas XPI " initial page in 
the Book of Kells the incomplete frame terminates in 
a human head at one end and two legs at the other. 
Another initial page in the same MS. — that of St. 
Mark's Gospel — has a zoomorphic frame, but the 
beast's head is holding a man between its jaws, whilst 
the man is tugging at the beast's tongue with his hand. 

Groups of four human figures arranged swastika 
fashion, interlaced and each grasping the limbs, wrists, 

^ A pin brooch ornamented with a human head, from Woodford 
River, Co. Cavan, is illustrated in Sir W. Wilde's Catal. of the Mus. 
i?./.A,p. 565. 


hair, or beard of one of the other figures, occur on 
crosses in Ireland at Kilkispeen, Co. Kilkenny ; Mon- 
asterboice, Co. Louth ; and Kells, Co. Meath ; and in 
Scotland on a recumbent monument at Meigle, Perth- 
shire. A human figure interlaced with a bird occurs 
in two instances on sculptured stones in Scotland, 
namely, at Monifieth, Forfarshire (now in the Edin- 
burgh Museum of Antiquities) ; and at Meigle, Perth- 


Leaf and plant motive decoration is entirely foreign 
to the spirit of purely Celtic Christian art, and whenever 
it occurs it is generally to be traced to Northumbrian 
influence. The Book of Kells and the Stockholm 
Gospels are the only Hiberno-Saxon illuminated MSS. 
in which any trace of foliage can be detected. There 
are panels of foliage on the Irish crosses at Kells, Co. 
Meath ; Monasterboice, Co. Louth ; and Clonmacnois, 
King's Co. In Wales there is an instance of foliage 
on the crosses at Penally, Pembrokeshire. In Scotland 
the only sculptured monuments with foliage upon 
them (excluding, of course, those in the Northumbrian 
districts of the south) are the erect cross-slabs at 
Hilton of Cadboll and Tarbet, Ross-shire (both now at 
Invergordon Castle) ; St. Vigeans, Forfarshire ; and 
Crieff, Perthshire ; on crosses at Camuston, Forfar- 
shire ; Dupplin, Perthshire ; and on a cross-shaft at 
St. Andrews, Fifeshire. 

The foliage may in all cases be traced back to the 
Classical vine, the well-known symbol of Christ. It 
is often much degraded by successive copying, and 
although the forms of the leaves are often altered 
beyond recognition the bunches of grapes can always 
be made out. 




We have already mentioned most of the figure- 
subjects to be found In the Hiberno-Saxon illuminated 
MSS., and on the Irish ecclesiastical metalwork. It 
remains therefore only now to take the sculptured 
monuments into consideration. 

It was in Ireland alone that a recognised cycle of 
scriptural figure-subjects was adopted for the decoration 
of the crosses and that in nearly all cases the ornament 
was relegated to a subordinate position. In Scotland 
and Wales, on the contrary, Scripture scenes are 
seldom represented on the sculptured monuments ; in 
Cornwall the only figure-subject which occurs on the 
crosses is the Crucifixion ; and in the Isle of Man the 
figure-subjects are mostly taken from the Pagan Norse 

The following table shows the Scriptural subjects 
on the sculptured monuments of Ireland, Scotland, 
Wales, and Cornwall, and the frequency with which 
they occur : — 





Old Testament— 

Adam and Eve . 



. — . 

Noah in the Ark 



Sacrifice of Isaac 



■ — ■ 

Three Children in Furnace . 


- — - 

Daniel in Den of Lions 



■ — 

David and Harp 



David and Lion . 



David and Goliath 



Jonah and Whale 


Ascent of Elijah 









New Testament — 

Virgin and Child . 




Adoration of Magi 




Flight into Egypt . 


Baptism of Christ . 


Miracle of Loaves and Fishes 



Raising of Lazarus 


Crucifixion .... 





Christ in Glory 




Last Judgment 




Annunciation .... 




Christ seized by the Jews 



Twelve Apostles 




Agnus Dei .... 



■ — 

Dextera Dei .... 




In addition to the above there are the following, 
which are sacred or ecclesiastical, but not, strictly 
speaking. Scriptural : — 

Ireland. Scotland. Wales. Cornwa 

Symbols of Four Evangelists 
Cherubim .... 
Angels .... 
Saints .... 
Oranti .... 

It appears, then, that the Scriptural subjects of most 
frequent occurrence in Ireland are the Crucifixion, 
Adam and Eve, the Sacrifice of Isaac, Daniel in the 
Lions' Den, and the scenes from the Life of David ; 
and in Scotland, the Crucifixion, Daniel in the Lions' 
Den, the Virgin and Child, and the symbols of the 
four Evangelists. 



The subjects common to both Ireland and Scotland 
are Adam and Eve, Noah (?), Sacrifice of Isaac, Daniel 
in the Lions' Den, David and the Harp, David and 
the Lion, Adoration of the Magi, Flight into Egypt, 
Miracle of Loaves and Fishes, Crucifixion, Christ in 
Glory, Agnus Dei, Angels. 

The subjects which occur in Ireland, but not in 
Scotland, are the Three Children in the Furnace, 
David and Goliath, Baptism of Christ, Resurrection, 
Last Judgment, Dextera Dei, Twelve Apostles. And 
those which occur in Scotland, but not in Ireland, are 
Ascent of Elijah, Raising of Lazarus, Jonah and the 
Whale, Annunciation, Salutation, Miracle of Healing 
the Blind, Christ and Mary Magdalene, Lazarus. 

Of the subjects on the early sculptured stones of 
Ireland and Scotland the following belong to the cycle 
of subjects found on the paintings in the Catacombs 
and the Sculptured Sarcophagi (a.d. 50 to 450) : — 

Adam and Eve. 

Sacrifice of Isaac. 
Three Children in the Fur- 
Ascent of Elijah. 

Daniel in the Lions' Den. 
Jonah and the Whale. 
Adoration of Magi. 
Miracle of Loaves and Fishes. 
Miracle of Healing the 

The following subjects belong to the Lombardo- 
Byzantine period (a.d. 700-1100): — 


Christ in Glory. 

Baptism of Christ. 

Last Judgment. 


Agnus Dei. 


Dextera Dei. 

Flight into Egypt. 

Twelve Apostles. 

Virgin and Child (apart from 

Symbols of the Four Evan 



Christ and Mary Magdalene. 



Thus the early Sculptured Stones and the Hiberno- 
Saxon MSS. of Great Britain, and the Carlovingian 
Ivories afford a connecting link between the older 
symbolism of the primitive Christianity of the Cata- 
comb period and the more strictly ecclesiastical art of 
mediaeval times. 

Quite apart from the fact that King David was a type 
of Christ, and that his pictures formed the illustrations 
of the Psalter, it is not surprising that he should have 
been an object of popular worship amongst the warlike 
and musical Celts, to one side of whose character his 
heroic deeds in rending the jaws of the lion and slay- 
ing the giant Goliath, would appeal as strongly as his 
talent as a harper would to the other. 

A small MS. Irish Psalter in the British Museum 
(Vit. F. i.)^ contains two very curious miniatures, one 
of David Playing the Harp and the other of David 
and Goliath.' The former is interesting, because I 
think it helps to explain the meaning of a figure sitting 
on the back of a beast and playing a harp,^ sculptured 
on one of the panels of the cross at Clonmacnoise. 
As I hold, this is intended for David ; and my reason 
for supposing this is, because the throne on which 
David is seated in the miniature in the Psalter is 
conventionally treated as a beast. 

I am not quite sure whether the boat with men in it, 
on the stone at Cossins, is intended for Noah's Ark or 
not, but a boat of just the same kind is represented on 

^ Westwood's Miniatures, pi. 5. 

^ In the miniature of David and Goliath in the Psalter David holds a 
sling in one hand and a beast-headed club in the other. The resem- 
blance between this club and the beast's-head symbol, which occurs on 
the Norrie's Law silver ornaments and on several of the early incised 
slabs in Scotland, may be only accidental, but it is worth noting- as a 
possible clue to the scriptural interpretation of the symbol. 

'^ O'Neil, pi. 24A. 


a carved wooden pillar at Olaf's Church, ^ Nesland, 
where it is associated with other Scriptural subjects, 
amongst others the creation of Eve, Samson and 
Delilah, and David and Goliath. In this case there 
can be little doubt but that the boat is intended for 
Noah's Ark, so that probably the boat at Cossins has 
the same meaning. 

The angels are cherubim, with four wings, and 
spirals where the wings join on to the body, represen- 
tations of which are to be seen on the stones at Eassie, 
Glamis, and elsewhere in Scotland. They do not occur 
on any of the sculptured crosses in Ireland ; but there 
are instances of angels or the symbols of the four 
evangelists treated in the same fashion in the St. Gall 
Gospels, Codex No. 51,- and on the Book Shrine of St. 
Molaise's Gospels,'^ in the Museum of the Royal Irish 
Academy in Dublin, and also on a bronze plaque^ of 
the Crucifixion, in the same collection. I have recently 
discovered a very curious instance of an angel of this 
kind, with three wings on a cross-slab, with interlaced- 
work, in St. David's Cathedral, given in Westwood's 
Lapidarium WallicE (pi. 63, fig. 4), but the wings and 
spirals only shown, and the head of the angel omitted. 

The pair of ecclesiastics, sometimes standing, some- 
times enthroned, sometimes kneeling, with a bird 
holding a circular disc in its mouth between them, is a 
subject common to the early sculptured stones of both 
Scotland'' and Ireland,'' but the exact meaning of it 

^ L. H. S. Dietrichscii, De Norskc Stavkirker, p. 362. 

^ C. Purton Cooper's "Appendix A to Report on Rynier's Foedera," 


'■' Archceologia, vol. xliii., p. 131. 

■* Westwood's Miniatxires, pi. 51. 

•^ As at Nig-g- and St. Vigeans. Dr. J. Anderson regards the Nigg- 
example as being intended for St. Paul and St. Anthony. 

^ As at Kells, Moone Abbey, Clonca. 


has yet to be ascertained if we are not to take the 
instance on the Ruthwell cross as an authoritative 
explanation of the whole. 

As I have already pointed out in my Rhind Lectures 
on Christian Symholism^ there is a nearer affinity 
between the subjects chosen to decorate the bases of 
the Irish crosses and the representations of hunting 
scenes, horsemen, chariots, etc., on the upright cross- 
slabs of the north-east of Scotland, than the more 
strictly Scriptural scenes on the shafts of the Irish 
crosses. The best examples illustrating this are to be 
seen on the bases of the crosses at Kells (Figs. 5 and 6), 
Monasterboice, Clonmacnois, Castle Dermot (Fig. 7), 
and Kilklispeen. 

The chariot on the Meigle slab, now lost, may be 
compared with the chariots to be seen on the shaft of 
the cross at Killamery, and on the bases of the crosses 
at Monasterboice, Kilklispeen, and in Kells churchyard; 
on the base of the cross in the street at Kells we have 
the eagle and fish, as on the " Drosten " stone at 
St. Vigeans,^ and as in the Book of Armagh ; and on 
the base of this same cross, and on the cross of 
Muredach at Monasterboice, centaurs occur, in some 
respects like those on the slabs at Aberlemno, Meigle, 
and Glamis. 

On the base of the Kilklispeen cross is portrayed 
a procession of ecclesiastics taking part in a most 
remarkable ceremony. On the south side of the base 
is to be seen a priest carrying a processional cross, 
and followed by a man leading a horse, on the back 
of which is laid the headless trunk of a man, with 

1 Another remarkable instance of the eagle ^md fish has recently been 
found on a stone with an Og^am inscription, at Latheron, near Keiss, 


two birds of prey, or carrion crows, perched on the 

On the north side of the base are two ecclesiastics 
on horseback, followed by two more in a chariot drawn 
by a pair of horses. 

On the east side are several beasts, birds, and a 

On the west side is a central figure, perhaps a bishop, 
with three ecclesiastics holding croziers on each side of 

These scenes can hardly be Scriptural ; and if they 
are not taken from the life of some saint, it is difficult 
to see what explanation remains to be suggested, 
except that an event of local importance is here com- 
memorated. The bases of the pillar-cross at Llandough 
and of the great wheel-cross at Margam, both in 
Glamorganshire, are the only ones with figures of 
horsemen upon them in Wales. 

The symbolism of the shafts of the Irish crosses is 
so strictly biblical that secular subjects may have been 
placed on the bases by way of contrast, to indicate the 
actual world or earth on which the cross stood repre- 
senting the spiritual world. The eagle and fish may 
personify the ocean, and the centaur the desert, for 
which we have the authority of the bestiaries and the 
legendary life of St. Anthony. 

The points of similarity between the ornamental 
patterns on the stones of Ireland and Scotland raise 
questions of too much intricacy to be dealt with here ; 
but it may be remarked that figure-sculpture forms the 
chief feature of the Irish crosses — geometrical, zoo- 
morphic, and foliageous designs being only as a rule 
applied to the decoration of the smaller panels on the 
sides of the shafts and to the rings connecting the 


arms. The upright cross-slabs of Scotland, more 
particularly those in Ross-shire, approach much more 
nearly in style — and therefore probably in age — to the 
illuminated pages of the Hiberno-Saxon MSS. of the 
best period, than do any of the Irish crosses. 

In conclusion, I consider the so-called Celtic style to 
be a local variety of the Lombardo-Byzantine style, 
from which the figure-subjects, the interlaced-work, 
the scrolls of foliage, and many of the strange real 
and fabulous creatures were apparently borrowed. 
The Lombardo-Byzantine style was introduced into this 
country after the Saxons had become Christians ; and 
being grafted upon the Pagan art of the Late-Celtic 
period, was developed in different ways in different 
parts of Great Britain. However, it in no way detracts 
from the artistic capacity of the Celt that he should 
have adapted certain decorative motives belonging to a 
foreign style instead of evolving them out of his own 
inner consciousness. Although his materials may not 
all have been of native origin, they were so skilfully 
made use of in combination with native designs, and 
developed with such exquisite taste, that the result was 
to produce an entirely original style, the like of which 
the world had never seen before. 


Abercorn, cross-shaft, 269 
Aberg-ele, horse-trapping-s, 95 
Aberlemno, erect cross-slab, 184 
Abernethy, 82 
Abing-don, pottery, 124 
Aboyne, armlet, 113 • 

— erect cross-slab, 184 
Achnabreac, spirals at, 50 
JEgea.n, spiral ornament in the, 54 
iEsica, 82 

— fibula, 152 

Alfrlston, horse-trappings, 94 

— enamelled harness-ring-, 134 
Alg-eria, penannular brooch, 225 
Ailing-ton, pottery, 124 
Alstonfield, Iron Ag-e iDurial at, 68 
Amber setting-s, 237 
Amerden, sword-sheath, 91 
Ammendola, sarcophagus of, 6 
Andrews, St., sarcophagus, 193 
Ang-els, 300 

Animals, fig-ures of, 146 

Antennae, swords with, 86 

Anthony, St., 302 

Anthropomorphic desig-ns, Late- 
Celtic, 144 

Anthropomorphs in Celtic Christian 
art, 294 

Apollinare in Classe, Sant', Ra- 
venna, 244 

Archdall, Castle, spirals at, 51 

Ardagh, chalice, 216, 235 

Ardakillen Crannog, 81 

— — fibula, 170 

Ardchattan, erect cross-slab, 184 
Ardoch, 82 

— horse-trappings, 95 
Aristotle, 2 
Armagh, bell, 199 

— Book of, 175 

— shrine of Book of, 208 

Armlets, enamelled, 134 

— Late-Celtic, 113 

— of glass, Late-Celtic, 141 
Arria and Paetus, 5 
Arras, armlet, 114 

— horse-trappings, 94 

— Iron Ag-e burials at, 63 

— mirror, 1 15 
Aryans, i 

Aspatria, sculptured cist, 55 
Assynt, jet necklace from, 42 
Aston Clinton, pottery, 124 
Athenry, dagger-sheath, 92 
Athlone, Crucifixion, 218 
Auchendolly, horse-trappings, 95 
Auchenbadie, armlet, 113 
Avebury, fibula, 106 
Axe-head of bronze with spirals, 52 

— sculptures, 55 
Aycliflfe, cross-shaft, 269 
Aylesford, bucket, 116, 145 

— fibula, 106 

— Iron Ag-e cemetery at, 70 

— pottery, 122 

— tankard, 1 16 

Ay ton Moor, cup-marks at, 57 

Backworth, pair of fibulse, 103 

— saucepan, 117 
Badony, Lower, bell, 198 
Baglan, cross-slab, 182 
Bakerhill, sculptured cist, 55 
Balcalk, jet necklace from, 42 
Ballinderry Crannog, 81 

comb, 128 

Balmaclellan, Late-Celtic finds at, 


— mirror, 115 

Balls, stone, of Bronze Age, 50 
Ballycostello, horse-trappings, 95 
Ballynaminton, horse-trappings, 95 



Bandag-Ing- patterns, 38 

Bangor, bell, 201 

Bapchild, horse-trappings, 94 

Barg-any House, sword-sheath, 91 

Barlaston, bowl, 58, 166 

Barlaston, Iron Age burial at, 68 

Barrington, enamelled bowl, 135 

Barrochan, cross, 192 

Bartlow Hills, enamelled vessel, 138 

Basketry, Late-Celtic, 143 

Battersea, shield, 92 

— sword-sheath, 91 
Beaded torques, 159 
Beads, Late-Celtic, 125 141 
Belhelvie, armlet, 113 
Bells, Celtic, 194 
Bell-shrines, 204 

Benty Grang-e, enamelled bowl, 134 

— Iron Ag'e burial at, 67 
Berkshire, triskele pendant, 121 
Bernaldby Moor, sculptured cist, 55 
Berru, helmet from, 12 
Beverley, Iron Ag-e burials at, 65 
Bibracte, enamels from, 138 
Big-bur}- Camp, 82 

Birdlip, bowl, 1 16 

— fibula, 106 

— Iron Age burials at, 69 

— mirror, 1 15 
Birnie, bell, 198 
Birrenswark, 82 

Birrenswark, horse-trappings, 95 
Biskra, penannular brooch, 225 
Black Hedon, cup-marks at, 57 
Blackshaw, spirals at, 50 
Blathmac, crozier, 208 

Boars' tusks from Arras, 64 
Bones, engraved, 150 
Bonework, Late-Celtic, 127 
Bonville, fibula, 106 
Book-shrines, 208 
Bowl, Barlaston, 166 

— Chesterton, 167 

— enamelled, Moklebust, 160 
Bowls, Late-Celtic, 116 

— with enamelled handles, 134 
Boxmoor, sword-sheath, 91 
Braddan, Kirk, wheel-cross, 188 
Braintree, pottery, 124 
Braughing-, enamelled vessel, 137 
Brazing, 132 

Breac Moedoc, 211 
Breaks in plait, 259 
Bride, Kirk, cross, 186 

Bridle-bit, enamelled. Rise, 133 

Birrenswark, 133 

Brig-hton, pottery, 125 
Brodie, erect cross-slab, 184 
Broig-hter, Limavady, find of gold 

ornaments, 109 
Brooch, penannular, 223 
Bronze Age burials, 23 

Celtic art in, 22 

Celts, 15 

chronolog-y of, 19 

patterns which survived into 

the Iron Age, 60 
Brough, 82 

Brough, fibula, 107, 154 
Brythons, 17 
Buckets, 145 

— Late-Celtic, 1 16 
Burg-head, 82 
Burials of Iron Ag-e, 65 
Bunrannoch, armlet, 113 
Byzantine art, 239, 244 

Caburn, Mount, Late-Celtic oppi- 

dum at, 74 
Caistor, enamelled bowl, 135 
Calderstones, spirals at, 50 
Camborne, cross-slab, 182 
Came Down, cujD-marks at, 57 
Camerot Muir, spirals at, 50 
Canterbury, chatelaine, 115 

— horse-trappings, 94 

Capel Garmon, fire-dogs, 118, 141 

Cappagh, bell, 199 

Carew, cross, 193 

Carham, sword-shcath, 91 

Carlingwark Loch, tankard, 116 

Carlsvvark Cavern, armlet, 114 

Carlton, inlaid object, 139 

Carnban, sculptured cist, 55 

Carn Bre, 16 

Carnwath, sculptured cist, 55 

Cashel, bell, 201 

Castell Nadolig, 82 

Casting- metal, Iron Ag-e, 130 

Castle Dermot, 192 

Castle Newe, armlet, 113 

weem at, 82 

— ■ — enamelled armlets, 134 
Castlethorpe, armlet, 1 14 
Catterdale, sword-sheath, 91 
Cauldrons of bronze, 86 
Caves inhabited by Brit-Welsh, 75, 



Celtse, 4 

Celtic tribes in Britain, 17 
Celt in classical sculpture, 5 
Celts of bronze, 41 
Centaur, 302 
Certosa situlee, 145 
Chad, St., Book of, 175 
Chains, 237 

— of gold and silver, Late-Celtic, 

Chalices, Celtic, 215 
Chalk drums from Folkton, 56 
Champlevt^ enamel, 136 
Characteristics of Celtic Christian 

style, 254 

— Late-Celtic style, 143 
Chariots, 301 

— wheels, 98 

— wheels from Arras, 64 
Chatelaines, Late-Celtic, 115 
Chedworth, Chi-Rho Monog-ram, 

Chequerwork, Late-Celtic, 160 
Cherubim, 300 
Chesterford, Great, Kimmeridge 

shale vessel, 128 
Chesters, Great, fibula, 107 
Chesterton, bowl, 167 

— enamelled bowl, 135 
Chevron patterns, 28 

Late-Celtic, 160 

Chi-Rho Monogram, 163 
Chorley, chains, 140 

— pair of fibulffi, 102 
Christianity in Britain, 162 
Chronology of Bronze Age, 19 
Circles, concentric, in Bronze Age, 


Circular knotwork, 273 

Cirencester, fibula, 107 

Cists of Bronze Age with sculp- 
ture, 55 

Cividale, Baptistery of Calistus, 258 

Claughton Moor, cup-marks at, 57 

Clemente, San, Rome, 244, 258 

Clogher, fibula, 105 

— bell, 199 
Cloisonn^ enamel, 136 
Clonmacnois, 190 

— cross-slabs, 180 

— crozier, 207 

— gold collar, in 

— pin, chains, 140 

— pin, 219 

Clooncunra, horse-trappings, 95 
Clova, horse-trappings, 95 
Clyde Crannog, comb, 128 
Coins, ancient British, 77 

— with Late-Celtic finds, 88 
Cookham, dagger-sheath, 92 
Colchester, Kimmeridge shale ves- 
sel, 128 

Commios, coins of, 78 

Corfe Castle, Kimmeridge shale 

vessel, 128 
Cowlam, fibula, 106 
Cochno, wheel symbol at, 59 
Coffey, George, theory on spiral 

ornament, 53 
Coffins, stone, 193 
Coilsfield, spirals at, 50 

— sculptured cist, 55 
Coinage, Celtic, 14 
Collar, Late-Celtic, 153 
Collars, Late-Celtic, 109 
Colours used by Celtic scribes, 233 
Columba's Psalter, shrine of, 209 
Compass-work in Late-Celtic art, 

Cong, cross, 213 
Coped stones, 182 
Copenhagen, reliquary, 211 
Coral settings, 237 
Cossins, erect cross-slab^ 184 
Courcelles-et-Montagne, 13 
Cowlam, armlet, 114 

— Iron Age burials at, 65 
Craigie Wood, sculptured cist, 55 
Craigy warren, pin, 108 
Crannogs, 81 

Craven Arms, chatelaine, 115 
Crawfordjohn, silver chains, 170 
Cremation in Bronze Age, 23 
Crichie, horse-trappings, 95 
Cricklade, fibula, 106 
Crosby Ravensworth, spoons, 120 
Cross, Celtic, evolution of, 186 

— processional, 213 
Cross-slabs, 181 

Crossthwaite, enamelled bowl, 134 
Croy, penannular brooch, 228 
Crozlers, Celtic, 206 
Cruciform breaks in plait, 262 

— patterns, 236 
Crucifixion plaques, 218 
Crystal settings, 237 
Cuerdale, penannular brooch, 227 
Culan, St., bell-shrine, 204 



Culbin Sands, armlet, 113 

Cumdachs, 208 

Cup-and-ring- sculpture at Ilkley,59 

Wooler, 59 

Kirkcudbright, 59 

Lochg-ilphead, 59 

— ■ — Kilmartin, 59 
Cup-marked stones, 57 

Danes' Graves, horse-trappings, 94 

— — Iron Age burials, 65 

pin, 107 

Dartmoor, 16 

Datchet, fibula, 106, 139 

Dates of Late-Celtic finds, 85 

Decorative motives in Bronze Age, 

Delphi sacked by the Gauls, 4 
Denmark, spiral ornament in, 51 
Deepdale Cave, 76, 81 

— looped wire armlets, 139 
Diagonal lines, use of, 60 

— setting-out lines, 255 
Dimma's Book, shrine of, 209 
Discs of repouss^ bronze, Late- 
Celtic, 121 

Dogmael's, St., 18 
Dogs, fire-, 141 

Late-Celtic, 118 

Dorchester, sword-sheath, 91 
Dornoch Links, cup-marks at, 57 
Dots, rows of, in decoration, 39 

— use of, 60 
Dowalton Crannog, Si 

— horse-trappings, 95 
Dowkerbottom Cave, fibula, 107 
Dowth, spirals at, 50 
Drinking-cup urns, 25 
Drumclifif, cross, 192 
Drumragh, bell, 199 
Dunfallancly, erect cross-slab, 184, 

Dunkeld, Little, bell, 198 
Dupplin, cross, 192 
Durrow, cross, 192 

— Book of, 169, 174 

— shrine of Book of, 208 
Dyce, erect cross-slab, 184 
Dysert, crozier, 208 

Eagle and fish, 301 
Ecclesiastics, 301 
Eclay, spirals at, 50 
Egypt, spiral ornament in, 54 

Eilean Finan, bell, 198 
Eliseg's Pillar, Valle Crucis, 193 
Elkstone, cup-marks at, 57 
Elveden, pottery, 124 

— tankard, 116 
Embleton, sword-sheath, 91 
Embsay, collar, 1 1 1 
Emlagh, horse-trappings, 95 
Enamelling, Iron Age, 133 

— process of, 135 

Enamels of Christian Celtic period, 

Engraved patterns. Iron Age, 132 
Eppillos, coins of, 78 
Essex, pottery, 122 
Evolution of Celtic cross, 186 

Farley Heath, 82 

— fibula, 107 
Farnell, cross-slab, 184 
Farr, erect cross-slab, 184 
Fens, the, horse-trappings, 94 
Fibula, evolution of, 99 

— from ^sica, 152 
Fibulae, enamelled, 134 

— disc-shaped, 154 

— worn in pairs with chain, 104 
Figure drawing in Celtic Christian 

art, 256 
Figure-subjects in Celtic Christian 

art, 296 
Filigree-work, 237 
Finds, Late-Celtic, dates of, 85 
— geographical distribution of, 

with coins, 88 

— of Iron Age, nature of, 61 
Fire-dogs, 141 

— Late-Cehic, 118 
Firth, spirals at, 50 
Flamboyant ornament, Late-Celtic, 

Flasby, sword-sheath, 91 
Flutes, Late-Celtic, 118 
Foliage in Celtic Christian art, 247, 


— in Late-Celtic art, 148 
Folkton, chalk drums from, 56 
Food-vessel urns, 25 

Ford West Field, sculptured cist, 55 
Forres, erect cross-slab, 184 
Forteviot, bell, 198 
Frampton, Chi-Rho Monogram, 



Galati, 4 

Gall, St., Gospels of, 177 

Galli, 4 

Garton, bell, 198 

Gaulcross, pin, 108 

Gauls, Cisalpine, 13 

Geographical distribution of Late- 
Celtic finds, 79 

Gheg-an Rock, comb, 128 

Giacomo, San, Venice, 245 

Gilton, mirror, 115 

Giorg"io, San, di Valpolicella, 244, 

Gladiator, the Dying-, 5 

Glamis, erect cross-slab, 184 

Glass, Late-Celtic, 125, 141 

— settings, 237 

Glastonbury Lake-Dwellings, 72,81 

— Lake Village, pottery, 142 
woodwork, 147, 161 

— Marsh Villag-e, pottery, 125 

woodwork, 126 

bowl, 117 

Glencolumbkille, spirals at, 51 
Glencotho, sword-sheath, 91 
Glenfahan, Ogam stone, 165 
Gobnet, St., stone of, 165 
Goidelic Celts, 15 

Goidels, 17 

Golden Grove, cross, 261 

— ■ — - cross-shaft, 282 

Golspie, erect cross-sla,b, 184 

Gorg-e-Meillet, helmet from, 12 

Gospels, 173 

Goulien, bell, 199 

Gourdon, chalice, 215 

Go van, sarcophagus, 193 

Grsekwyl, 13 

Grang-e of Conan, weem at, 82 

armlet, 113 

Greenloan, cup. marks at, 57 
Greenwich, enamelled bowl, 135 
Grimspound, 16 
Grimthorpe, Iron Ag-e burial at, 65 

— sword-sheath, 91 

Guthrie Castle, bell-shrine, 204 

Hagbourne Hill, horse-trapping's, 95 

Late-Celtic finds at, 76 

Hallristning-ar, 60 
Hallstatt, cemetery of, 7 

— situlfE, 145 

— type, objects of, found in Great 
Britain, 86 

Hambledown Hill, 82 

Ham Hill, 82 

Hamdon Hill, horse-trapping-s, 95 

Late-Celtic finds at, 76 

Hammer-headed pins, 108 
Hammersmith, fibula, 106 
Hand-grasping- fig-ures, 294 
Harness-mountings, 98 

— enamelled, 134 
Harness-ring-s, enamelled, 134 

— Late-Celtic, 96 
Harray, broch of, 82 
Haughton - le - Skerne, sword - 

sheath, 91 
Hay Hill, fire-dog's, 118 
Head ornaments for horses, 98 
Hecatseus, 2 
Helen's, St., Chi-Rho Monogram, 

Helmet from Berru, 12 

— from Gorg-e-Meillet, 12 
Helmets, Late-Celtic, 93 
Henshole, horse-trappings, 95 
Herodotus, 2 

Hexag-on patterns, 36 
High Crosses, 188 
Highfield Pits, pottery, 125 
Hilton of Cadboll, erect cross-slab, 

Hinksey, N. , dag-g-er-sheath, 92 
Hitchin, pottery, 124 
Hod Hill, 82 

fibula, 106 

sword-sheath, 91 

Hollows Tower, spirals at, 50 
Hook-and-disc ornaments, Late- 
Celtic, 121 
Horns with terminal knobs, 148 
Horse's head ornaments, 98 
Horse-trapping-s from Arras, 64 

— Late-Celtic, 94 

Hounslow, fig-ures of animals, 146 
Hunsbury, fibula, 106 

— Late-Celtic oppidum at, 73 

— pottery, 125 

— sword-sheath, 132 

— sword-sheath, 91, 97 

— triskele pendant, 121 
Hunterston, brooch, 231 
Hyndford Crannog-, 81 
collar, 112 

Iberians, 15 

Ickling-ham, sword-sheath, 91 



Ilkley, cup-and-ring- sculpture, 59 

— rock-sculptures at, 57 
Incense-cup urns, 25 
Inchbra3^ock, erect cross-slab, 184 
Ing^oe, cup-marks at, 57 
Inhumation in Bronze Age, 23 
Inishkeel, bell, 203 

Ink used by Celtic scribes, 232 
lona, cross, 192 
Insh, bell, 198 
Interlaced-work, 246 

— evolution of, 257 
Inverg-owrie, erect cross-slab, 1S4 
Ireland, spoons, 121 
Italo-Greek objects found at Ayles- 

ford, 71 
Iron Age finds, 61 

Jet necklaces, 42 

Jezerine, chains, 140 

Just, St., Chi-Rho Monogram, 163 

Kells, Book of, 174 
Kells, cross, 192, 224, 233 

— crozier, 207 

— shrine of Book of, 208 
Kelko Cave, 76, 81 
Kent's Cavern, 76, 81 

— — pottery, 125 
Keshkerrigan, bowl, 117 
Keston, 16 

Kettleburn, broch of, 82 
Key-patterns, 278 
Kilbroney, bell, 198 
Kilburn, cup-marks at, 57 
Kildalton, cross, 192 

Kildare, Co. ,penannular brooch, 228 
Kilfenora, cross, 192 
Kilkeeran, horse-trapping's, 95 
Kilklispeen, cross, 192 
Killarney, crozier, 207 

— gold lunula from, 40 

Killeen Cormac, Ogam stone, 165 
Killing Hill, spirals at, 51 
Kilmainham, bell, 198 
Kilmartin, cup-and-ring sculpture, 


— sculptured cist, 55 
Kilmichael Glassary, bell-shrine, 


Kilnsea, fibula, 107 

Kimmerid^e shale objects, Late- 
Celtic, 128 

Kingsholm, pottery, 125 

Kingsholm, triskele pendant, 121 
King-'s Mountain, spirals at, 50 
Kingston Down, enamelled bowl, 


Kirkby Thore, fibula, 107 

Kirkcudbright, cup-and-ring- sculp- 
ture, 59 

Kirkhead Cave, 75, 81 

Kirkly Thore, 82 

Kirkmadrine, Chi-Rho Monogram, 


Kirk Whelpington, cup-marks at, 57 

Kirriemuir, horse-trappings, 95 

Kit's Coty House, pottery, 124 

Knobs on horns of beasts, 148 

Knockmany, spiriils at, 51 

Knot, spiral, 270 

Knots used in Celtic art, 265 

Knotwork, circular, 273 

— triangular, 276 

Lactin's, St., Arm, shrine of, 210 
Lagore Crannog-, 81 

beads, 125 

comb, 128 

— pin, 108 
Lake-dwellings, 81 
Lake-dwellings at Glastonbury, 72 
Lanivet, coped stone, 182 
Lashing- patterns, 38 
Late-Celtic finds, dates of, 85 

— — geog-raphical distribution of, 

with coins, 88 

— style, 143 

La Tfcjne, oppidura of, 9 
Latheron, cag-le and fish, 301 
Lattice-work patterns, 34 
Leather-work, 237 
Leicester, horse-trappings, 94 
Lilburn Hill, spirals at, 50 
Limavady, g-old chains, 139 
Limavady, g-old collar, 109, 153 
Lincoln, sword-sheath, 91 
Lindisfarne Book, 177 
Linlithg-owshire, enamelled patera, 

Lismore, crozier, 207 
Lisnacrog-hera Crannog-, 81 
sword-sheath, 148 

— sword-sheath, 91 
Llanbadarn Fawr, cross, 193 
Llanbedr, spirals at, 50 
Llanfair, spoons, 121 





Llandoug-h, cross, 193 
Llandyssyl, collar, 1 1 1 
Llang-enydd, cross-shaft, 270 
Llangystenyn, bell, 198 
Llang-wynodl, bell, 200 
Llanrhyddlad, bell, 198 
Llantwit Major, cross, 186, 261 
Lochar Moss, bowl, 116 

collar, 1 1 1 

Lochg-ilphead, cup-and-ring- sculp- 
ture, 59 
Lochlee Crannog-, 81 
woodwork, 127 

— horse-trappingfs, 95 
Lonan, Kirk, wheel-cross, 188 
London, enamelled harness-ring-, 


— fibula, 106 

— horse-trappings, 94 

— spoons, 120 

— sword-sheath, 91 
Looped wire work, 139 
Lorrha, bell, 201 
Loug-hcrew, spirals at, 50 
Lough Erne, reliquary, 210 
Lough Lene Castle, bell, 201 
Loug-hnashade, trumpet, 118 
Loug-h Ravel Crannog, beads, 125 
Lozenge patterns, ^2 

Late-Celtic, 160 

Lullingstone, enamelled bowl, 135 

Lunulse, gold, 41 

Lumphanan, stone ball from, 50 

MacDurnan, Gospels, 175 
MacRegol, Gospels of, 175 
Madoes, St., erect cross-slab, 183 
Maen Achwyfan, cross, 193 
Maes Mynach, cross-shaft, 270 
Magny Lambert, 12 
Maiden Castle, 82 
Maiden Stone, erect cross-slab, 184 
Maltbeck, enamelled vessel, 138 
Malton, fibula, 107 
Malton, New, 82 
Marlborough, bucket, 116 
Mancha, La, spirals at, 50 
Manchan, St., shrine of, 212 
Manuscripts, illuminated, 173 
Margam, cross, 186 
Marlborough bucket, 144 
Marne, La, cemeteries of, 11 
Materials for study of Celtic art in 
Bronze Ag-e, 22 

Materials for study of Celtic art in 
Iron Age, 90 

Materials for study of Celtic Chris- 
tian art, 172 

Materials used by Celtic sculptors, 

Maug-hanby, cup-marks ^lt, 57 

— spiriils at, 50 
Maughold, Kirk, cross, 1S6 
Meigle, coped stone, 182 

— erect cross-slab, 184 
Melfort, jet necklace from, 42 
Metallurg-y, Iron Age, 130 
Metals used by Christian Celts, 

Metal-work, Christian Celtic, 194 
Methods employed by Celtic sculp- 
tors, 233 
Methods employed by Celtic metal- 
workers, 234 
Mevagh, wheel and symbol at, 59 
Middleby, horse-trappings, 95 
Middleton Common, Iron Age burial 

at, 67 
Middleton Moor, enamelled bowl, 


jet necklace from, 42 

Migvie, erect cross-slab, 184 
Mirrors, Late-Celtic, 115 
Mirror, Trelan Bahow, 131 
Mogue, St., bell-shrine, 204 
Moklebust, enamelled bowl, 160 
Molaise's Gospels, shrine of, 209 
Monasterboice, cross, 190, 224 
Monifieth, erect cross-slab, 184 
Monogram, Chi-Rho, 163 
Moone Abbey, cross, 192 
Mont Beuvray, enamels from, 138 

pottery, 141 

Monuments, sculptured, of Chris- 
tian period, 180 
Monymusk, reliquary, 210 
Moresby, pin, 108 
Moreton Hall, sword-sheath, 91 
Morwick, spirals at, 50 
Motives, decorative, in Bronze 

Age, 23 
Motives used in Christian Celtic 

art, 242 
Mount Batten, Iron Age cemetery 

at, 70 
Mount Bures, fire-dogs, 118 
Mount Caburn, pottery, 125 
Mountings for harness, 98 



Mowroad, collar, 1 1 1 
Mura, St., bell-shrine, 204 
Mycensean art, 57 

Nancy, chalice, 215 

Navan Rath, fibula, 106 

Necklaces of jet, 42 

Needham Market, enamelled bowl, 

Neolithic dwelling-s, 16 
Neuadd Siarman, cross, 190 
Nevern, cross, 191 
Newcastle, chains, 140 
Newgrang-e tumulus, 43 
Newry, armlet, 113 
Niello, 235 

Nig-g-, erect cross-slab, 1S2 
Norrie's Law, leaf-shaped plates, 


pin, loS 

Northfield, Late-Celtic village, 82 
Norton, enamelled fibula, 134 

— enamelled harness-rings, 134 

— horse-trappings, 94 

Ogam-inscribed stones, 165 
Okstrow, broch of, 82 

— tankard, 116 

Old Parks, tumulus, 56 
Orange, Triumphal Arch of, 5 
Orchomenos, spiml ornament iit, 54 
Origin of early Christian art, 238 
Ornovasso, chains, 140 
Orton Scar, penannular brooch, 227 
Over-Haddon, enamelled bowl, 134 
Oxford, enamelled bowl, 135 

Pails, Late-Celtic, 116 

P and Q Celts, 18 

Panels, ornament arranged in, 255 

Papil, erect cross-slab, 184 

Parisi, 12 

Patera, enamelled, 137 

Patrick's Will, St., bell-shrine, 204 

Patrick's, St., Gospels, shrine of, 

Penally, cross, 193 
Penannular brooch, 223 
Pen Arthur, cross-slab, 18 r 
Penbryn, spoons, 121 
Penmachno, Chi-Rho Monogram, 

Penmon, cross, 185 

Pens used by Celtic scribes, 232 
Pentre Poeth, Og^am stone, 165 
Perdeswell, collar, 1 1 1 
Perfection of details of Celtic orna- 
ment, 256 
Perforated metal plates, 236 
Perg'amos, 5 

Perth, penannular brooch, 231 
Philostratus, 133 
Pins, Celtic, 219 

— hammer-headed, 108 

— Late-Celtic, 107 
Pitalpin, armlet, 113 
Pitkelloney, armlet, 113 

— enamelled armlets, 134 
Plait, cruciform breaks in, 262 

— horizontal and vertical breaks 
in, 259 _ 

Plaques with Crucifixion, 218 

Plato, 2 

Plaitwork, 259 

Plunton Castle, armlet, 113 

Polden Hill, enamelled harness- 

— fibula, 106 

horse-trapping's, 95 

Pol dc Leon, St., bell, 201 
Polybius, 3 
Poole's Cave, 76, 81 
Port-Blanc, dolmen, 39 
Portland, Isle of, collar, iii 
Pottery, Late Celtic, 121, 142 

— painted, Mont Beuvray, 141 

— sepulchral, in Bronze Age, 24 
Prassede, San, Rome, 258 
Presles, enamels from, 138 
Prickwillow, saucepan, 117 
Processes, technical, in Celtic Chris- 
tian art, 232 

Iron Age, 129 

Psalter of St. John's Colleg-e, Cam- 
bridge, 178 

— Vesp. A. i., 178 

— Vit. F. xi., 178 
Psalters, 176 

Pyrmont, enamelled vessel, 138 
Pytheas, 3 

Rathconrath, wooden bowl, 127 
Ravenna, screen, 245 
Reask, inscribed stone, 166 
Rectilinear patterns in Late 

art, 159 
Redlands, scu 

patterns in Late-Celt 
ilptured cist, 55 



Relic shrines, 210 

Repouss^ metal - work, Iron Age, 

Rhayader, gold armlet, 170 
Ribchester, 82 

— fibula, 107 

Ricemarchus, Psalter of, 176 
Ring-ham Low, fibula, 106 
Rings for harness, 96 

Rise, horse-trappings, 94 
Risingham, 82 

— enamelled fibula, 134 

— fibula, 107 
Riveting, Iron Age, 132 
Rock-sculpture in Scandinavia, 60 
Rodenbach, 13 

Rogart, penannular brooch, 228 
Rome taken by the Gauls, 4 
Rosemarkie, erect cross-slab, 184, 

Rossie Priory, erect cross-slab, 184 
Rotherley, 82 
Round towers, 196 

Sadberge, sword-sheath, 91 
Safety-pin type of fibula, 100 
Saham Toney, cruciform harness- 
mounting, 98 
enamelled harness-mounting, 


horse-trappings, 94 

Saltire patterns, 35 

Sandy, pottery, 124 

Sarcophagi, 193 

Saucepans, Kelto-Roman, 117 

Scandinavia, rock-sculpture in, 60 

— spiral ornament in, 51 
Scandinavian features in Celtic art, 

Scattery Island, bell, 198 
Scribes, pens, ink, etc., used by, 232 
Scripture subjects in Celtic art, 296 
Sculptured monuments of Christian 

period, 180 
Sculptured stone at Newgrange, 49 
Sculptured stones, Late-Celtic, 128 
Sculpture on Bronze Age cists, 55 
Seafield Tower, armlet, 113 
Seal-box, enamelled, 134 
Seamill Fort, 82 

triskele pendant, 121 

Sepulchral remains of Iron Age, 63 
Seskin, sculptured cist, 55 
Sesto-Calende, 9 

Settings of coral, amber, glass, and 

crystal, 237 
Shading, kinds of, in Late-Celtic 

art, 156 
Shandwick, erect cross-slab, 184 
Sheen, cup-marks at, 57 
Shields, Late-Celtic, 92 

— South, horse-trappings, 94 
Shoebury, pottery, 124 
Shrines of bells, 204 

— of books, 208 

— of relics, 210 
Silchester, 82 

— enamelled stand, 134 

— fibula, 107, 154 
Silian, cross-shaft, 270 
Silures, 18 

Situlffi, Hallstatt, 145 

— Late-Celtic, 116 

— of bronze, 86 

Skaill, penannular brooch, 227 
Sliabh na Calliaghe, 43 

bonework, 127 

— engraved bones, 150 

Soldering, 132 

Somme-Bionne, 13 

Southill, Chi-Rho Monogram, 163 

Southwark, dagger-sheath, 92 

Spear-head of Bronze Age, 39 

Spiral, closely coiled, 60 

— knot, 270 

• — ornament at Newgrange, 48 

at Orchomenos, 54 

in Bronze Age, 50 

in Egypt, 54^ 

in Scandinavia, 51 

Spirals in Celtic Christian art, 284 

— Late-Celtic, 154 

— from Book of Durrow, 169 
Spoons, Late-Celtic, 119 
Stamfordbury, fire-dogs, 118 
Stamford Hill, armlet, 114 
■ fibula, 106 

mirror, 1 15 

Stand, enamelled, 134 
Standon, enamelled vessel, 138 
Stanhope, armlet, 113 
Stanton, cup-marks at, 57 
Stanwick, horse-trappings, 94 

— Late-Celtic finds at, 75 

— sword-sheath, 91 
Step-patterns, 277 

— in Late-Celtic art, 160 
Stitchell, collar, 11 1 



Stival, bell, 200 
Stockholm, Gospels, 175 
Stones, sculptured, Late-Celtic, 128 
Storr, penannular brooch, 227 
Stowe Missal, cover of, 170 

shrine of, 209 

Strathfillan, bell, 201 
Strokestown Crannog', 81 

• fibula, 170 

Strypes, spirals at, 50 
Style, Late-Celtic, 143 
Swastika anthropomorphs, 294 

— curved, at Ilkley, 58 

on Barlaston bowl, 58 

, 60 

— designs, Late-Celtic, 154 
Sweden, bronze axe-head with 

spirals from, 52 
Sword, Hallstatt, 8 
Sword-sheath from La T^ne, 1 1 

— Hunsbury, 132 
Sword-sheaths, Late Celtic, 91 
Swords with antennae, 86 
Symbols used in Bronze Age, 22 

Tankard, Late-Celtic, 142, 151 
Tankards, Late-Celtic, 116 
Tara brooch, 229 
chains, 140 

— horse-trappings, 95 

— horse's head ornament, 98 
Tassilo, chalice of, 215 
Tayfield, jet necklace, 42 
Technical processes. Iron Age, 


in Celtic Christian art, 232 

Termonfechin, cross, 192 
Textile patterns. Iron Age, 129 
Thames, river, helmet, 93 

Late-Celtic finds in, 77 

Thirst House Cave, armlet, 114 

chatelaine, 115 

fibula, 107 

Thor's Cave, 76, 81 

flutes, 118 

Tillycoultry, sculptured cist, 55 
Tincommios, coins of, 78 
Trawsfynydd, tankard, 142, 151 
Treceiri, 82 

— triskele pendant, 121 
Trelan Bahow, mirror, 115, 131 

Iron Age burials at, 69 

Trenoweth, collar, 1 1 1 
Triangle patterns, 31 

Triangular knotwork, 276 
Triskele anthropomorphs, 294 

— designs, Late-Celtic, 154 

— pendants, Late Celtic, 121 
Torques, beaded, 159 
Torrish, jet necklace from, 42 
Torrs, helmet, 93 
Torwoodlee, broch of, 82 

— horse-trappings, 95 
Towers, round, 196 
Towie, stone ball from, 50 
Trumpet, Late-Celtic, 118 
Tuam, cross, 210 

Tully lease, cross-slab, 181 
Tumulus at Newgrang-e, 34 

— at Old Parks, 56 

Turoe, Late-Celtic sculpture, 128 
Tweed, river, Late-Celtic finds in, 

Tyne, fibula, 107 

— river, Late-Celtic finds in, 77 

Uffizi Museum, cruciform harness- 
mounting, 98 
enamelled harness-mounting, 

Ulbster, erect cross-slab, 184, 276 
Lhistan, urn from, 38 
Urns, cinerary, 24 
Urquhart, pin, 108 

Victoria Cave at Settle, 76, 81 

fibula, 107, 154 

Vigeans, St., erect cross-slab, 184 
Verica, coins of, 78 

Walmer, fibula, 105 

— spoon, 121 

Wandsworth, dagger-sheath, 92 
Warden, mirror, 115 

Warden, Old, Kimmeridg-e shale 

vessel, 128 
Warren, Folkestone, fibula, 102 
Warton, sword-sheath, 91 
Water Eaton, fibula, 106 

— — sword-sheath, 91 
Watsch situlse, 145 
Wellingborough, 82 
Westhall, horse-trappings, 94 

— enamelled harness-ring-s, 134 
Weston, spoons, 121, 147 
Weymouth, pottery, 124 
Wheel-crosses, 188 




Wheel symbol in Bronze Age, 59 
Whitechurch, pottery, 124 
Winding--band patterns, 57 
Witham, river, dag-ger-sheath, 92 

Late-Celtic finds in, 77 

shield, 92, 139 

Whithorn, Chi-Rho Monogram, 

— cross-shaft, 270 
Woodwork, Late-Celtic, 126 
Woodwray, erect cross-slab, 184 

Wooler, cup-and-ring sculpture, 59 

Wraxhall, collar, 1 1 1 

Wrought metal -work. Iron Age, 

Wykeham Moor, cup-marks at, 57 

Zoomorphic designs, Late-Celtic, 

Zoomorphs in Celtic Christian art, 

249, 287